PBRM1 BIOMARKERS PREDICTIVE OF ANTI-IMMUNE

CHECKPOINT RESPONSE

Cross-Reference to Related Applications

This application claims the benefit of U.S. Provisional Application No. 62/445,094, filed on 11 January 2017; the entire contents of said application are incorporated herein in their entirety by this reference.

Background of the Invention

Immune checkpoint inhibitors, including monoclonal antibodies targeting programmed cell death- 1 (PD-1) and its ligand PD-Ll, have yielded durable clinical benefit for patients with a range of tumor types, including metastatic melanoma, non-small cell lung cancer (NSCLC), and bladder cancer (Topalian et al. (2015) Cancer cell 27:450-461). Renal cell carcinoma (RCC) has been known to be immunotherapy -responsive for more than 20 years, with high-dose interleukin-2 yielding durable complete responses in a small minority of patient. Recently, immune checkpoint blockade has been shown to be remarkably effective in RCC, yielding superior rates of objective response compared to everolimus (Motzer et al. (2015) N. Engl. J. Med. 373 : 1803-1813). However, responses only occurred in about a quarter of patients, and immunohistochemistry for PD-Ll was not predictive of treatment response (Choueiri et al. (2016) Clin. Cancer Res. 22:5461-5471; Motzer et al. (2015), supra), making identification of pre-treatment predictors of patient benefit a clinical priority.

Studies featuring somatic genetic analysis of patients receiving immune checkpoint therapies for metastatic melanoma, non-small cell lung cancer, and colorectal cancer have demonstrated that pre-treatment tumor mutational load, neoantigen burden, microsatellite instability, gene expression signatures, and neoantigen clonality can influence likelihood of response (Hugo et al. (2016) Cell 165:35-44; Le et al. (2015) N. Engl. J. Med. 372:2509- 2520; McGranahan et al. (2016) Science 351 : 1463-1469; Rizvi et al. (2015) Science 348: 124-128; Snyder et al. (2014) N. Engl. J. Med. 371 :2189-2199; and Van Allen et al. (2015) Science 350:207-211). These studies support the concept that highly mutated tumors generate tumor-specific antigens (neoantigens) that mediate a strong immune response to cancer cells after the administration of immune checkpoint therapies that disrupt immunosuppression in the tumor microenvironment. In contrast to melanoma, non-small cell lung cancer, and microsatellite-unstable colorectal cancer, which commonly harbor more than 10 to 400 mutations per megabase (Mb), clear cell renal cell carcinoma (ccRCC) has an average of 1.1 nonsynonymous mutations/Mb, without significant outliers (Cancer Genome Atlas Research (2013) Nature 499:43-49), while ranking among the highest across multiple tumor types in cytolytic activity (Rooney et al. (2015) Cell 160:48-61), immune infiltration score, and T cell infiltration score (§enbabaoglu et al. (2016) Genome Biol. 17:231), suggesting that distinct molecular mechanisms other than mutational burden or neoantigen burden may underlie its T-cell enriched microenvironment and responsiveness to immune checkpoint therapy. RCC is also characterized by frequent alterations in von Hippel Lindau protein (VHL), a tumor suppressor that regulates the transcription factor hypoxia inducible factor la (HIF1 A) central to controlling angiogenesis. Alterations in SWItch/Sucrose Non-Fermentable (SWI/SNF) chromatin remodeling complex are also common. SWI/SNF subunits commonly mutated in ccRCC include polybromo 1 (PBRMl), AT-Rich Interaction Domain 1 A (ARIDl A), and Transcription activator BRG1 (SMARCA4) are also common. Other commonly mutated genes included the histone deubiquitinase BRCA1 Associated Protein 1 (BAP1), and the histone methyltransferase SET domain containing 2 (SETD2). The genes encoding VHL, PBRMl, BAPl, and SETD2 are all clustered in the small arm of chromosome 3 (chr3p), and arm-level deletions of chr3p are exceedingly common in ccRCC (>90% of samples; TCGA (2013) Nature 499:43-49). While the relationship between these DNA-level alterations affecting chromatin remodeling, angiogenesis, and response to hypoxia and the enrichment in immune cell infiltration in ccRCC is still not fully understood, experimental studies aiming to characterize the functional impact of PBRMl loss have identified upregulation of the interleukin-6-mediated signaling pathway as one effect of re-expressing PBRMl in PBRMl -deficient RCC cell lines (Chowdhury et al. (2016) PLoS One 11 :e0153718).

In clinical studies of patients receiving anti-PD-1 therapy for metastatic RCC, whole genome microarray characterization of pre-treatment tumors from 11 patients revealed that nonresponders had higher expression of genes related to cell metabolism and solute transport, while responders overexpressed immune markers (Ascierto et al. (2016) Cancer Immunol Res. 4:726-733). Germline variants in STAT3, a transcription factor associated with immune function, have also previously been linked to response to immunotherapy with high-dose interferon (Eto et al. (2013) Eur. Urol. 63 :745-752). However, no study has yet examined pre-treatment tumor whole exome and whole transcriptome sequencing with matched germline whole exome sequencing in well-annotated cohorts of renal cell carcinoma patients treated with immune checkpoint inhibitor therapy (e.g., anti-PDl therapy used to treat metastatic RCC) to discover alterations in specific genes,

transcriptional profiles, and immunological features that may predict response to immune checkpoint therapy. Accordingly, there remains a great need in the art to identify biomarkers predictive of response to immune checkpoint therapy for improved clinical stratification and enhanced understanding of the mechanism of these drugs. Summary of the Invention

The present invention is based, at least in part, on the discovery that loss-of-function mutation of PBRMl is a highly specific biomarker for prediction of clinical outcomes (e.g., improved clinical outcomes such as tumor shrinkage and prolonged survival) in renal cell carcinoma patients treated with immune checkpoint therapies, such as those comprising an anti-PD-1 therapeutic (e.g., PD-1 blocking antibody).

In one aspect, a method of identifying the likelihood of a cancer in a subject to be responsive to an immune checkpoint therapy, the method comprising a) obtaining or providing a subject sample from a patient having cancer; b) measuring the amount or activity of at least one biomarker listed in Table 1 in the subject sample; and c) comparing said amount or activity of the at least one biomarker listed in Table 1 in a control sample, wherein the absence of or a significantly decreased amount or activity of the at least one biomarker listed in Table 1 in the subject sample and/or the presence of or a significantly increased amount or activity of the at least one biomarker listed in Table 1 having a loss of function mutation in the subject sample, relative to the control sample identifies the cancer as being more likely to be responsive to the immune checkpoint therapy; and wherein the presence of or a significantly increased amount or activity of the at least one biomarker listed in Table 1 in the subject sample and/or the absence of or a decreased amount or activity of the at least one biomarker listed in Table 1 having a loss of function mutation in the subject sample, relative to the control sample identifies the cancer as being less likely to be responsive to the immune checkpoint therapy, is provided.

In another aspect, a method of identifying the likelihood of a cancer in a subject to be responsive to immune checkpoint therapy, the method comprising a) obtaining or providing a subject sample from a patient having cancer, wherein the sample comprises nucleic acid molecules from the subject; b) determining the copy number of at least one biomarker listed in Table 1 in the subject sample; and c) comparing said copy number to that of a control sample, wherein a decreased copy number of the at least one biomarker listed in Table 1 in the in the subject sample and/or an increased copy number of the at least one biomarker listed in Table 1 having a loss of function mutation in the subject sample, relative to the control sample identifies the cancer as being more likely to be responsive to the immune checkpoint therapy; and wherein a wild type or increased copy number of the biomarker in the subject sample and/or or a decreased copy number of the at least one biomarker listed in Table 1 having a loss of function mutation in the sample relative to the control sample identifies the cancer as being less likely to be responsive to the immune checkpoint therapy, is provided.

Numerous embodiments are further provided that can be applied to any aspect of the present invention and/or combined with any other embodiment described herein. For example, in one embodiment, the method provided herein further comprises

recommending, prescribing, or administering the immune checkpoint therapy if the cancer is determined likely to be responsive to the immune checkpoint therapy or administering an anti-cancer therapy other than the immune checkpoint therapy if the cancer is determined be less likely to be responsive to the immune checkpoint therapy. The anti-cancer therapy may be, for example, selected from the group consisting of targeted therapy, chemotherapy, radiation therapy, and/or hormonal therapy. In another embodiment, the control sample described herein is determined from a cancerous or non-cancerous sample from either the patient or a member of the same species to which the patient belongs. In still another embodiment, the control sample is a cancerous or non-cancerous sample from the patient obtained from an earlier point in time than the patient sample. In yet another embodiment, the control sample is obtained before the patient has received immune checkpoint therapy and the patient sample is obtained after the patient has received immune checkpoint therapy. In another embodiment, the control sample described herein comprises cells or does not comprise cells. In still another embodiment, the control sample comprises cancer cells known to be responsive or non-responsive to the immune checkpoint therapy.

In another aspect, a method of assessing the efficacy of an agent for treating a cancer in a subject that is unlikely to be responsive to an immune checkpoint therapy, comprising a) detecting in a first subject sample and maintained in the presence of the agent the amount or activity of at least one biomarker listed in Table 1 ; b) detecting the amount or activity of the at least one biomarker listed in Table 1 in a second subject sample and maintained in the absence of the test compound; and c) comparing the amount or activity of the at least one biomarker listed in Table 1 from steps a) and b), wherein the presence of or a significantly increased amount or activity of the at least one biomarker listed in Table 1 in the first subject sample and/or the absence of or a decreased amount or activity of the at least one biomarker listed in Table 1 having a loss of function mutation in the first subject sample, relative to at least one subsequent subject sample, indicates that the agent treats the cancer in the subject, is provided.

In another aspect, a method of assessing the efficacy of an agent for treating a cancer in a subject or prognosing progression of a cancer in a subject, comprising a) detecting in a subject sample at a first point in time the amount or activity of at least one biomarker listed in Table 1; b) repeating step a) during at least one subsequent point in time after administration of the agent; and c) comparing the expression and/or activity detected in steps a) and b), wherein the presence of or a significantly increased amount or activity of the at least one biomarker listed in Table 1 in the first subject sample and/or the absence of or a decreased amount or activity of the at least one biomarker listed in Table 1 having a loss of function mutation in the first subject sample, relative to at least one subsequent subject sample, indicates that the cancer is unlikely to progress or that the agent treats the cancer in the subject, is provided. In one embodiment, between the first point in time and the subsequent point in time, the subject has undergone treatment, completed treatment, and/or is in remission for the cancer. In another embodiment, the first and/or at least one subsequent sample is selected from the group consisting of ex vivo and in vivo samples. In still another embodiment, the first and/or at least one subsequent sample is obtained from an animal model of the cancer. In yet another embodiment, the first and/or at least one subsequent sample is a portion of a single sample or pooled samples obtained from the subject.

In another aspect, a cell-based assay for screening for agents that have a cytotoxic or cytostatic effect on a cancer cell that is unresponsive to an immune checkpoint therapy comprising, contacting the cancer cell with a test agent, and determining the ability of the test agent to decrease the amount or activity of at least one biomarker listed in Table 1 in the subject sample and/or increase the amount or activity of the at least one biomarker listed in Table 1 having a loss of function mutation, is provided. In one embodiment, the step of contacting occurs in vivo, ex vivo, or in vitro. In another embodiment, the subject sample and/or the control sample has not been contacted with a renal cell cancer treatment or inhibitor of an immune checkpoint. In still another embodiment, the subject has not been administered a renal cell cancer treatment or inhibitor of an immune checkpoint. In yet another embodiment, the method or the cell-based assay provided herein further comprises recommending, prescribing, or administering at least one additional anti-cancer therapeutic agent. In another embodiment, the at least one additional anti-cancer therapeutic agent is nivolumab and/or an anti-PBRM-1 therapeutic agent.

As described above, numerous embodiments are contemplated for any aspect of the present invention described herein. For example, in one embodiment, the subject sample is selected from the group consisting of serum, whole blood, plasma, urine, cells, cell lines, and biopsies. In another embodiment, the amount of the at least one biomarker listed in Table 1 is detected using a reagent which specifically binds with the protein. For example, the reagent may be selected from the group consisting of an antibody, an antibody derivative, and an antibody fragment. In still another embodiment, the at least one biomarker listed in Table 1 is assessed by detecting the presence in the sample of a transcribed polynucleotide or portion thereof. For example, the transcribed polynucleotide may be an mRNA or a cDNA. The transcribed polynucleotide can be detected by identifying a nucleic acid that anneals with the biomarker nucleic acid, or a portion thereof, under stringent hybridization conditions. In yet another embodiment, the step of detecting further comprises amplifying the transcribed polynucleotide. In another embodiment, the at least one biomarker listed in Table 1 is human PBRM-1, or a fragment thereof. In still another embodiment, the immune checkpoint therapy described herein comprises at least one antibody selected from the group consisting of anti-PD-1 antibodies, anti-CTLA-4 antibodies, anti-PD-Ll antibodies, anti-PD-L2 antibodies, and combinations thereof. For example, the immune checkpoint therapy may comprise nivolumab. In yet another embodiment, the likelihood of the cancer in the subject to be responsive to immune checkpoint therapy is the likelihood of at least one criteria selected from the group consisting of cellular proliferation, tumor burden, m-stage, metastasis, progressive disease, clinical benefit rate, survival until mortality, pathological complete response, semi- quantitative measures of pathologic response, clinical complete remission, clinical partial remission, clinical stable disease, recurrence-free survival, metastasis free survival, disease free survival, circulating tumor cell decrease, circulating marker response, and RECIST criteria. In another embodiment, the cancer is a solid tumor. In still another embodiment, the cancer is a renal cell cancer. In yet another embodiment, the renal cell cancer is a clear cell renal cell cancer (ccRcc). In another embodiment, the clear cell renal cell cancer is a metastatic clear cell renal cell carcinoma (mRCC). In still another embodiment, the subject described herein is a mammal. In yet another embodiment, the mammal is an animal model of cancer. In another embodiment, the mammal is a human.

Brief Description of Figures

Figure 1 includes 5 panels, identified as panels A, B, C, D, and E, which show the cohort consolidation and clinical characteristics of the training cohort. Panels A and B summarize the clinical cohort investigated unless otherwise indicated, such as at Panel D. Generally, of the 91 patients who received anti-PDl monotherapy (nivolumab) as part of CA209-009, 56 had available pre-treatment tumor for whole exome sequencing. After quality control, 34 pre-treatment tumors were processed through standardized analytical pipelines and included in the final analysis cohort (Panel B). Sixteen samples (the leftmost column) were excluded for low sample purity (including patients who had early death on treatment) (Panel C). Patient were classified into clinical response groups based on objective tumor response RECIST classifications (complete response: CR, partial response: PR, stable disease: SD, or progressive disease: PD) (CITE: RECIST) as well as duration of progression-free survival (PFS) (time from starting immune checkpoint therapy to experiencing objective tumor growth). "Extreme responders" had CR or PR by RECIST or SD with objective tumor shrinkage lasting >6 months) while "extreme progressors" experienced PD by RECIST with PFS < 3 months). A third group called "intermediate benefit" or "stable disease" had responses to therapy intermediate between the extreme responders and extreme progressors, based on a combination of objective tumor response by RECIST and duration of progression-free survival. Patients' overall survival (OS) following initiation therapy (in years) vs. PFS (in years) and PFS vs. decrease in tumor burden are shown in Panels C and D. One patient with early minor tumor growth followed by sustained tumor shrinkage was classified as an extreme-responder despite short PFS (see Figure 2).

Figure 2 includes 4 panels, identified as panels A, B, C, and D, which show the patient response classifications described in Figure 1. One patient (5 50) had early tumor growth (likely pseudoprogression) in the setting of overall response to therapy followed by sustained tumor response and was classified as an extreme responder despite disease progression by RECIST criteria prior to 6 months. The results shown in Panels A and B versus those of Panels C and D correspond to the clinical cohort described in Panels A and D, respectively, of Figure 1.

Figure 3 includes 4 panels, identified as panels A, B, C, and D, which compare the patient survival probability vs. different clinical characteristics, including different groups receiving different dosages of treatment (Panel A), different sexes (Panel B), pre-treatment tumor immunohistochemical staining for the PD-1 ligand PD-Ll (Panel C), and response by RECIST criteria (Panel D). Kaplan-Meier analyses showed that baseline clinical characteristics, including pre-treatment PD-Ll immunohistochemistry (Panel C), did not influence overall survival. Objective tumor response by RECIST criteria was strongly associated with overall survival (p= .00027). Two patients who did not receive staging scans (RECIST not evaluable) following commencement of anti-PDl therapy were excluded from further analyses.

Figures 4 includes 5 panels, identified as panels A, B, C, D, and E, which show whole exome features of the training cohort (N=41). Overall number of detected mutations per sample (all changes to the DNA sequence of a gene) and nonsynonymous mutations (mutations that change the amino acid sequence of the resulting protein encoded by a gene) per sample were similar for patients classified as extreme progressors, extreme responders, or intermediate benefit (Panel A). The ratio of clonal to subclonal mutations was not associated with clinical benefit. Nonsynonymous mutational burden, mutations in gene commonly mutated in clear-cell renal cell carcinoma, estimated tumor purity by

ABSOLUTE (Carter et al. (202) Nat. Biotechnol. 30:413-421), and outcomes with immune checkpoint blockade are shown in a stacked CoMut plot (Panel B). The five shown genes were selected as the intersection between significantly mutated genes in TCGA clear-cell renal cell carcinoma and 7 genes significantly mutated by MutSigCV (Lawrence et al. (2013) Nature 499:214-218) in this cohort (see Table 2C). Truncating mutations in PBRM1 were significantly more common in extreme responders (8/9) vs. extreme progressors (3/12) (p = 0.0037; q = 0.026; Pearson's chi-squared, FDR over 7 genes significantly mutated by MutSigCV) (Panel C). Dashed red line indicates p < 0.01. Genes in black were significantly mutated across the entire cohort by MutSigCV, while genes in grey were mutated at lower levels. Patients with truncating alterations in PBRM1 had prolonged overall survival compared to those without truncating PBRM1 mutations (p = 0.042; Cox proportional hazards) (Panel D). Three patients with truncating alterations in PBRM1 who were "extreme progressors" due to early tumor growth on anti-PDl monotherapy had longer-than-expected overall survival (9 97: PFS 1.2 months, OS 28.6+ months and 13 96: PFS 1.2 months; OS 19.1+ months), with duration of overall survival being unevaluable in a third due to censoring (5 18: PFS 1.4 months, OS 3.6+ months) (Panel E).

Figure 5 show genes significantly mutated in extreme responders vs. extreme progressors. Of all 2,285 genes containing at least 1 nonsynonymous mutation in the training cohort, PBRM1 was the only gene mutated significantly more frequently in extreme responders vs. extreme progressors (8/9 extreme responders vs. 4/12 extreme progressors, p = 0.011; Pearson's chi-squared) prior to correcting for multiple hypothesis testing. Genes in black were significantly mutated across the entire training cohort according to MutSigCV, while genes in grey were not. Dashed red line indicates p < 0.01.

Figure 6 shows that patients with truncating mutations in PBRM1 had objective decreases in tumor burden and prolonged overall survival on immune checkpoint monotherapy.

Figure 7 includes 4 panels, identified as panels A, B, C, and D, which show the association between PBRM1 alterations and clinical benefit from immune checkpoint therapies in a validation cohort of patients with clear-cell renal cell carcinoma treated with monoclonal antibodies targeting PD-1 and PD-L1, either alone or in combination with monoclonal antibodies targeting the immune checkpoint cytotoxic T lymphocyte-associated protein 4 (CTLA-4). A clinical cohort of 41 patients treated with immune checkpoint therapy for metastatic renal cell carcinoma was narrowed to 28 patients in the final validation cohort (Panel A). Patients were stratified into extreme responder, extreme progressor, and intermediate benefit groups using the same definitions as in the training cohort (Panel B). Truncating alterations in PBRM1 were significantly more frequent in patients with extreme response to immune checkpoint monotherapy compared to those experiencing extreme progression (8/13 vs. 1/7) (p = 0.043; Pearson's chi-squared) (Panel C). Truncating alterations in PBRM1 frequently occurred in the context of heterozygous deletion of chromosome 3p, though 2 patients with frameshift alterations in PBRM1 who were copy-neutral at chromosome 3p also experienced extreme response (Panel D).

Figure 8 includes 4 panels, identified as panels A, B, C, and D, which show the gene set enrichment analysis (GSEA) (Subramanian et al. (2005) Proc. Natl. Acad. Sci. 102: 15545-15550) of gene sets significantly enriched in untreated ccRCC tumors from the TCGA in tumors with truncating alterations in PBRM1 versus those without. GSEA showed that PBRM1 -truncated tumors had significantly decreased infiltration of macrophages (Panel A), TH1 cells (Panel B), TH2 cells (Panel C), and T cells (Panel D).

Figure 9 includes 2 panels, identified as panels A and B, which show that expression of the immune checkpoints PD-Ll and PD-1 as well as CD8A and IFNG were significantly lower in PBRM1 -mutant compared to PBRMl-wildtype ccRCC (TCGA). No significant differences were noted in levels of expression of immune checkpoints (CTLA4, PDCDl : encoding PD-1, CD274: encoding PD-Ll, PDCD1LG2: encoding the PD-1 ligand PD-L2). Markers of cytolytic activity (GZMA, PRFl), interferon gamma (IFNy), or CD 8 T cells (CD8A) were noted between tumors with truncating mutations in PBRM1 (blue) versus those without (yellow) in pre-treatment patient samples (Panel A). However, analysis in a larger set of TCGA clear-cell RCC reveals significantly lower expression of CD8A (p=0.0093), IFNG (p= 0.00105), PD-L2 (p=0.0173), and PD-1 (p=0.0165) in PBRM1 -mutant tumors.

Figure 10 shows a lollipop plot summarizing PBRMl mutations described in the

Examples.

Figure 11 summarizes the clinical characteristics of RCC CA209-009 patients (N=91).

Figure 12 compares training cohort (CA209009) with validation cohort (DFCI + MSKCC). Mutations and copy number alterations in B2M, HLA, and other antigen presentation machinery were rare and did not segregate by response status.

Figure 13 includes 4 panels, identified as panels A, B, C, and D, which describe cohort consolidation and clinical characteristics of the discovery cohort. Panel A shows sample inclusion/exclusion criteria and computational workflow. Panel B shows clinical stratification by degree of objective change in tumor burden (y-axis) and duration of progression-free survival (x-axis). One patient (RCC 99) is not shown due to lack of tumor response data. *Patient RCC 50 was classified as clinical benefit despite PFS<6 months because there was continued tumor shrinkage after an initial period of minor tumor progression (see Figure 15). Panel C shows the mutation burden in the discovery cohort by response group. Panel D shows the ratio of subclonal to clonal mutations, as estimated by ABSOLUTE, by response group, ns = not significant. Abbreviations: CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease. Figure 14 includes 2 panels, identified as panels A and B, which show that clinical characteristics do not differ significantly between samples that passed and

failed whole exome sequencing in the discovery cohort (N = 35 pass, N = 20 fail). Panel A shows a distribution of the best RECIST scores of patients whose samples passed and failed sequencing. CR = complete response; PR = partial response; SD = stable disease; PD = progressive disease; E = not evaluable. Panel B show the overall survival (OS) and progression-free survival (PFS) distribution between patients with samples that passed sequencing and samples that failed, measured in days from anti-PD-1 treatment initiation.

Figure 15 includes 2 panels, identified as panels A and B, which shows spider plots of change in tumor burden for discovery cohort (N = 35). Panel A shows a spider plot showing change in target tumor size in the discovery cohort over time. Shading of lines corresponds to best response by RECIST: CR = complete response (purple), PR = partial response (pink), SD = stable disease (light green), PD = progressive disease (dark green). * Patient RCC 50 was classified as clinical benefit despite early (prior to 6 months) minor increase in tumor size (likely pseudo-progression), as this was followed by sustained tumor shrinkage. Patient RCC 99 is not shown due to early clinical disease progression and lack of re-staging scans after baseline. Panel B shows a spider plot shaded by response group in this study.

Figure 16 shows that pre-treatment clinical covariates did not predict response to immune checkpoint therapy. Dose of immune checkpoint therapy, patient sex, and PD-L1 immunohistochemical staining did not predict patient overall survival following anti-PD-1 therapy (p>0.05, log-rank test).

Figure 17 includes 2 panels, identified as panels A and B, which show alterations in HLA alleles and antigen presentation machinery in the discovery and validation cohorts. Mutations and copy number alterations in discovery cohorts (panel A) (N=35) and the validation cohort (panel B) (N=41; only tumors from the MSKCC and DFCI patients in the validation cohort (41 out of 69 total validation cohort patients) had raw sequencing data available for these analyses) are shown. One clinical benefit patient in the validation cohort had a heterozygous TAP1 nonsense mutation, while two B2M mutations occurred in the no clinical benefit cohort, one missense and one nonsense.

Figure 18 includes 4 panels, identified as panels A, B, C, and D, which show that tumor genome features in the discovery cohort reveals a correlation between PBRM1 LOF mutations and clinical benefit from anti-PD-1 therapy. Panel A shows mutations in the discovery cohort. Patients are ordered by response category, with tumor mutation burden in decreasing order within each response category. Shown are the genes that were recurrently mutated at a significant frequency, as assessed by MutSig2CV analysis. CNA = copy number alteration. Panel B shows enrichment of truncating mutations in tumors from patients in the CB vs. NCB groups. The top dashed line denotes q<0.1 (Fisher's exact test). Mutations in genes above the lower black dotted line are enriched in tumors of patients with CB from anti-PD-1 therapy and mutations in genes below the line are enriched in tumors of patients with NCB. Panel C shows a Kaplan-Meier curve comparing overall survival of patients treated with anti-PD-1 therapy whose tumors did or did not harbor LOF mutations in PBRM1. See also Figure 19 for a Kaplan-Meier curve comparing progression-free survival of these patients. Panel D shows a spider plot showing objective decrease in tumor burden in PBRM1-LOF vs. R5R 7 -intact tumors. Three patients with early progression on anti-PD-1 therapy and truncating mutations m PBRMl (darkest shading) had long and/or censored OS.

Figure 19 shows a Kaplan-Meier curve of discovery cohort patient progression-free survival by PBRM1 mutation status. PBRM1 truncating alterations are associated with increased progression-free survival following anti-PD-1 therapy (p=0.029; log-rank test).

Figure 20 includes 4 panels, identified as panels A, B, C, and D, which show that PBRM1 LOF mutations correlate with clinical benefit in a validation cohort of

ccRCC patients treated with immune checkpoint inhibitors. Panel A shows selection of the validation cohort. Panel B shows clinical outcomes in the validation cohort. Ten patients without post-treatment restaging scans (eight with clinical PD, two with SD, and one with PR) as well as 14 patients with targeted panel sequencing are not shown. Panel C shows the proportion of tumors harboring PBRM1 LOF mutations in patients in the CB vs. NCB groups. Error bars are S.E. *Fisher's exact p<0.05. Panel D shows truncating alterations in PBRM1 and response to anti-PD-(L)l therapies by sample. Shaded boxes indicate samples with truncating mutations in PBRM1, while light shading denotes samples without PBRM1 truncating mutations. Missense LOF denotes a missense mutation detected by targeted sequencing that was confirmed to be LOF by PBRM1 immunohistochemistry.

Figure 21 shows a Kaplan-Meier curve of combined discovery and validation cohort patient progression-free survival by PBRM1 LOF mutation status, stratified by therapy line. PBRM1 truncating alterations in patients who received anti-PD-(L) 1 therapy in a setting other than first-line (N = 81) were associated with increased progression-free survival (p = 0.0087, log-rank test). This association was not observed in patients who received immune checkpoint blockade as first-line therapy (N = 17).

Figure 22 includes 2 panels, identified as panels A and B, which show PBRMl mutational status in ccRCC influences immune gene expression. Panel A shows the results of GSEA performed on PB AF-deficient (A704BAF 180-/- and A704BAF 180wt, BRGl -/-) vs. PBAF-proficient (A704BAF180wt) kidney cancer cell lines using both Hallmark and corresponding Founder gene sets. GSEA enrichment plot shown for the KEGG cytokine- cytokine receptor interaction gene set in A704BAF180-/- vs. A704BAF180wt (PBRMl null vs. wildtype). The enrichment plot is similar for A704BAF180wt, BRGl-/- vs.

A704BAF180wt (BRGl null vs. wildtype); see Table 61. Panel B shows the results of GSEA also performed on RNA-seq from pre-treatment tumors in the discovery and validation cohorts of this study (n = 18 PBRM1-LOF vs. n = 14 PBRMl-mtact) using the Hallmark gene sets. Enrichment plots show increased expression of the hypoxia and IL6/JAK-STAT3 gene sets in the PBRMl -LOF tumors.

Figure 23 includes 3 panels, identified as panels A, B, and C, which show expression of immune genes and PBRMl in three independent ccRCC cohorts by

PBRMl mutation status. Panel A shows expression of immune checkpoints and immune cell markers in TCGA clear-cell renal cell carcinoma between i?RM/-loss-of-function (LOF) (N=104) and PBRMl -intact (N=288) tumors. Immune inhibitory ligands, including PDCDI, PDCD1LG2, LAG3, TIGIT, and VTCNI are significantly upregulated in PBRMl - intact versus PBRM1-LOF tumors (*q<0.05, **q<0.01). Panel B shows differential immune gene expression analysis in Sato et al. (N=73 i^SR Z-intact vs. N=19 PBRM- LOF) shows significant upregulation of VTCNI in PBRMl-mtact tumors (*p<0.05, **p<0.005). Panel C demonstrates that in N=32 patient tumors, no immune genes were significantly differentially expressed, although PBRMl-LOF tumors trended towards lower expression of most checkpoints (*p<0.05, **p<0.005). All three cohorts show significantly lower expression of PBRMl in PBRMl-LOF tumors compared to PBRMl -intact tumors (p=0.0027, 0.048, and 0.022, respectively), while tumors with non-truncating mutations in PBRMl more closely resembled the PBRMl -intact expression phenotype.

Figure 24 show immune gene expression in TCGA KIRC by VHL mutation status.

The presence or absence of truncating mutations in VHL did not correlate with expression levels of immune inhibitory ligands or other immune cell markers. Note that for every figure containing a histogram, the bars from left to right for each discrete measurement correspond to the figure boxes from top to bottom in the figure legend as indicated.

Detailed Description of the Invention

It has been determined herein that PBRM1 is a highly specific biomarker for predicted clinical outcome in cancer patients (e.g., renal cell carcinoma patients) receiving anti-immune checkpoint-based therapy (e.g., anti-PDl/PD-Ll agents alone or in combination with other anti-cancer therapeutics). Accordingly, the present invention relates, in part, to methods for stratifying patients and predicting response of a cancer in a subject to immune checkpoint therapy based upon a determination and analysis of mutations, described herein, of biomarkers, compared to a control. In addition, such analyses can be used in order to provide useful anti-immune checkpoint treatment regimens (e.g., based on predictions of clinical response, subject survival or relapse, timing of adjuvant or neoadjuvant treatment, etc.).

I. Definitions

The articles "a" and "an" are used herein to refer to one or to more than one (i.e. to at least one) of the grammatical object of the article. By way of example, "an element" means one element or more than one element.

The term "altered amount" or "altered level" refers to increased or decreased copy number (e.g., germline and/or somatic) of a biomarker nucleic acid, e.g., increased or decreased expression level in a cancer sample, as compared to the expression level or copy number of the biomarker nucleic acid in a control sample. The term "altered amount" of a biomarker also includes an increased or decreased protein level of a biomarker protein in a sample, e.g., a cancer sample, as compared to the corresponding protein level in a normal, control sample. Furthermore, an altered amount of a biomarker protein may be determined by detecting posttranslational modification such as methylation status of the marker, which may affect the expression or activity of the biomarker protein.

The amount of a biomarker in a subject is "significantly" higher or lower than the normal amount of the biomarker, if the amount of the biomarker is greater or less, respectively, than the normal level by an amount greater than the standard error of the assay employed to assess amount, and preferably at least 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, 150%, 200%, 300%, 350%, 400%, 500%, 600%, 700%, 800%, 900%, 1000% or than that amount. Such "significance" can be assessed from any desired or known point of comparison, such as a particular post-treatment versus pre-treatment biomarker measurement ratio (e.g., 1-fold, 1.1-fold, 1.2-fold, 1.3-fold, 1.4-fold, 1.5-fold, and the like) or a particular pre-treatment serum biomarker protein measurement (e.g., 2,500 pg/ml, 2,750 pg/ml, 3,000 pg/ml, 3, 175 pg/ml, 3,250 pg/ml, 3,500 pg/ml, and the like).

Alternately, the amount of the biomarker in the subject can be considered "significantly" higher or lower than the normal amount if the amount is at least about two, and preferably at least about three, four, or five times, higher or lower, respectively, than the normal amount of the biomarker. Such "significance" can also be applied to any other measured parameter described herein, such as for expression, inhibition, cytotoxicity, cell growth, and the like.

The term "altered level of expression" of a biomarker refers to an expression level or copy number of the biomarker in a test sample, e.g., a sample derived from a patient suffering from cancer, that is greater or less than the standard error of the assay employed to assess expression or copy number, and is preferably at least twice, and more preferably three, four, five or ten or more times the expression level or copy number of the biomarker in a control sample (e.g., sample from a healthy subjects not having the associated disease) and preferably, the average expression level or copy number of the biomarker in several control samples. The altered level of expression is greater or less than the standard error of the assay employed to assess expression or copy number, and is preferably at least 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, 150%, 200%, 300%, 350%, 400%, 500%, 600%, 700%, 800%, 900%, 1000% or more times the expression level or copy number of the biomarker in a control sample (e.g., sample from a healthy subjects not having the associated disease) and preferably, the average expression level or copy number of the biomarker in several control samples.

The term "altered activity" of a biomarker refers to an activity of the biomarker which is increased or decreased in a disease state, e.g., in a cancer sample, as compared to the activity of the biomarker in a normal, control sample. Altered activity of the biomarker may be the result of, for example, altered expression of the biomarker, altered protein level of the biomarker, altered structure of the biomarker, or, e.g., an altered interaction with other proteins involved in the same or different pathway as the biomarker or altered interaction with transcriptional activators or inhibitors. The term "altered structure" of a biomarker refers to the presence of mutations or allelic variants within a biomarker nucleic acid or protein, e.g., mutations which affect expression or activity of the biomarker nucleic acid or protein, as compared to the normal or wild-type gene or protein. For example, mutations include, but are not limited to substitutions, deletions, or addition mutations. Mutations may be present in the coding or non-coding region of the biomarker nucleic acid.

The term "PBRMl" refers to protein Polybromo-1, which is a subunit of ATP- dependent chromatin-remodeling complexes. PBRMl functions in the regulation of gene expression as a constituent of the evolutionary-conserved SWI/S F chromatin remodeling complexes (Euskirchen et al. (2012) J. Biol. Chem. 287:30897-30905). Beside BRD7 and BAF200, PBRMl is one of the unique components of the SWI/SNF-B complex, also known as polybromo/BRGl -associated factors (or PBAF), absent in the SWI/S F-A (BAF) complex (Xue et al. (2000) Proc Natl Acad Sci USA. 97: 13015-13020; Brownlee et al. (2012) Biochem Soc Trans. 40:364-369). On that account, and because it contains bromodomains known to mediate binding to acetylated histones, PBRMl has been postulated to target the PBAF complex to specific chromatin sites, therefore providing the functional selectivity for the complex (Xue et al. (2000), supra; Lemon et al. (2001) Nature 414:924-928; Brownlee et al. (2012), supra). Although direct evidence for PBRMl involvement is lacking, SWI/SNF complexes have also been shown to play a role in DNA damage response (Park et al. (2006) EMBO J.25:3986-3997). In vivo studies have shown that PBRMl deletion leads to embryonic lethality in mice, where PBRMl is required for mammalian cardiac chamber maturation and coronary vessel formation (Wang et al. (2004) Genes Dev. 18:3106-3116; Huang et al. (2008) Dev Biol. 319:258-266). PBRMl mutations are most predominant in renal cell carcinomas (RCCs) and have been detected in over 40% of cases, placing PBRMl second (after VHL) on the list of most frequently mutated genes in this cancer (Varela et al. (2011) Nature 469:539-542; Hakimi et al. (2013) Eur Urol. 63 :848-854; Pena-Llopis et al. (2012) Nat Genet. 44:751-759; Pawlowski et al. (2013) Int J Cancer. 132:E11-E17). PBRMl mutations have also been found in a smaller group of breast and pancreatic cancers (Xia et al. (2008) Cancer Res. 68: 1667-1674; Shain et al. (2012) Proc Natl Acad Sci t7&4.109:E252-E259; Numata et al. (2013) Int J Oncol. 42:403- 410). PBRMl mutations are more common in patients with advanced disease stage (Hakimi et al. (2013), supra), and loss of PBRMl protein expression has been associated with advanced tumour stage, low differentiation grade and worse patient outcome (Pawlowski et al. (2013), supra). In another study, no correlation between PBRMl status and tumour grade was found (Pena-Llopis et al. (2012), supra). Although PBRMl -mutant tumours are associated with better prognosis than BAP 1 -mutant tumours, tumours mutated for both PBRMl and BAPl exhibit the greatest aggressiveness (Kapur et al. (2013) Lancet Oncol. 14: 159-167). PBRMl is ubiquitously expressed during mouse embryonic development (Wang et al. (2004), supra) and has been detected in various human tissues including pancreas, kidney, skeletal muscle, liver, lung, placenta, brain, heart, intestine, ovaries, testis, prostate, thymus and spleen (Xue et al. (2000), supra; Horikawa and Barrett (2002) DNA Seq. 13 :211-215).

PBRMl protein localises to the nucleus of cells (Nicolas and Goodwin (1996) Gene

175:233-240). As a component of the PBAF chromatin-remodelling complex, it associates with chromatin (Thompson (2009) Biochimie. 91 :309-319), and has been reported to confer the localisation of PBAF complex to the kinetochores of mitotic chromosomes (Xue et al. (2000), supra). Human PBRMl gene encodes a 1582 amino acid protein, also referred to as BAF180. Six bromodomains (BD1-6), known to recognize acetylated lysine residues and frequently found in chromatin-associated proteins, constitute the N-terminal half of PBRMl (e.g., six BD domains at amino acid residue no. 44-156, 182-284, 383-484, 519- 622, 658-762, and 775-882 of SEQ ID NO:2). The C-terminal half of PBRMl contains two bromo-adjacent homology (BAH) domains (BAHl and BAH2, e.g., at amino acid residue no. 957-1049 and 1130-1248 of SE ID NO:2), present in some proteins involved in transcription regulation. High mobility group (HMG) domain is located close to the C- terminus of PBRMl (e.g., amino acid residue no.1328-1377 of SEQ ID NO:2). HMG domains are found in a number of factors regulating DNA-dependent processes where HMG domains often mediate interactions with DNA.

The term "PBRMl" is intended to include fragments, variants (e.g., allelic variants), and derivatives thereof. Representative human PBRMl cDNA and human PBRMl protein sequences are well-known in the art and are publicly available from the National Center for Biotechnology Information (NCBI). For example, two different human PBRMl isoforms are known. Human PBRMl transcript variant 2 (NM 181042.4) represents the longest transcript. Human PBRMl transcript variant 1 (NM 018313.4, having a CDS from the 115-4863 nucleotide residue of SEQ ID NO: l) differs in the 5' UTR and uses an alternate exon and splice site in the 3' coding region, thus encoding a distinct protein sequence (NP_060783.3, as SEQ ID NO:2) of the same length as the isoform (NP_851385.1) encoded by variant 2. Nucleic acid and polypeptide sequences of PBRMl orthologs in organisms other than humans are well known and include, for example, chimpanzee PBRMl (XM 009445611.2 and XP_009443886.1, XM_009445608.2 and

XP 009443883.1, XM 009445602.2 and XP 009443877.1, XM 016941258.1 and

XP_016796747.1 XM_016941256.1 and XP_016796745.1, XM_016941249.1 and XP_016796738.1 XM_016941260.1 and XP_016796749.1, XM_016941253.1 and XP_016796742.1 XM_016941250.1 and XP_016796739.1, XM_016941261.1 and XP_016796750.1 XM_009445605.2 and XP_009443880.1, XM_016941252.1 and XP_016796741.1 XM_009445603.2 and XP_009443878.1, XM_016941263.1 and XP_016796752.1 XM_016941262.1 and XP_016796751.1, XM_009445604.2 and XP_009443879.1 XM_016941251.1 and XP_016796740.1, XM_016941257.1 and XP_016796746.1 XM_016941255.1 and XP_016796744.1, XM_016941254.1 and XP_016796743.1 XM_016941265.1 and XP_016796754.1, XM_016941264.1 and XP_016796753.1 XM_016941248.1 and XP_016796737.1, XM_009445617.2 and XP 009443892.1 XM_009445616.2 and XP_009443891.1, XM_009445619.2 and XP_009443894.1 XM_009445615.2 and XP_009443890.1, XM_009445618.2 and XP_009443893.1, and XM_016941266.1 and XP_016796755.1), rhesus monkey PBRMl (XM_015130736.1 and XP_014986222.1, XM_015130739.1 and XP_014986225.1, XM 015130737.1 and XP 014986223.1, XM 015130740.1 and XP 014986226.1,

XM_015130727.1 and XP_014986213.1 XM_015130726.1 and XP_014986212.1 XM_015130728.1 and XP_014986214.1 XM_015130743.1 and XP_014986229.1 XM_015130731.1 and XP_014986217.1 XM_015130745.1 and XP_014986231.1 XM_015130741.1 and XP_014986227.1 XM_015130734.1 and XP_014986220.1 XM_015130744.1 and XP_014986230.1 XM_015130748.1 and XP_014986234.1 XM_015130746.1 and XP_014986232.1 XM_015130742.1 and XP_014986228.1 XM_015130747.1 and XP_014986233.1 XM_015130730.1 and XP_014986216.1 XM_015130732.1 and XP 014986218.1 XM_015130733.1 and XP_014986219.1 XM 015130735.1 and XP 014986221.1 XM 015130738.1 and XP 014986224.1 and

XM_015130725.1 and XP_014986211.1), dog PBRMl (XM_005632441.2 and

XP_005632498.1, XM_014121868.1 and XP_013977343.1, XM_005632451.2 and XP_005632508.1, XM_014121867.1 and XP_013977342.1, XM_005632440.2 and XP_005632497.1, XM_005632446.2 and XP_005632503.1, XM_533797.5 and

XP 533797.4, XM 005632442.2 and XP 005632499.1, XM 005632439.2 and XP_005632496.1, XM_014121869.1 and XP_013977344.1, XM_005632448.1 and XP_005632505.1, XM_005632449.1 and XP_005632506.1, XM_005632452.1 and XP_005632509.1, XM_005632445.1 and XP_005632502.1, XM_005632450.1 and XP_005632507.1 , XM_005632453.1 and XP_005632510.1, XM_014121870.1 and XP_013977345.1, XM_005632443.1 and XP_005632500.1, XM_005632444.1 and XP_005632501.1, and XM_005632447.2 and XP_005632504.1), cow PBRMl

(XM_005222983.3 and XP_005223040.1, XM_005222979.3 and XP_005223036.1, XM_015459550.1 and XP_015315036.1, XM_015459551.1 and XP_015315037.1, XM_015459548.1 and XP_015315034.1, XM_010817826.1 and XP_010816128.1, XM_010817829.1 and XP 010816131.1, XM_010817830.1 and XP_010816132.1, XM_010817823.1 and XP_010816125.1, XM_010817824.2 and XP_010816126.1, XM_010817819.2 and XP_010816121.1, XM_010817827.2 and XP_010816129.1, XM_010817828.2 and XP_010816130.1, XM_010817817.2 and XP 010816119.1, and XM_010817818.2 and XP_010816120.1), mouse PBRM1 ( M_001081251.1 and

P_001074720.1), chicken PBRM1 ( M_205165.1 and P_990496.1), tropical clawed frog PBRMl (XM_018090224.1 and XP_017945713.1), zebrafish PBRM1

(XM_009305786.2 and XP_009304061.1, XM_009305785.2 and XP_009304060.1, and XM_009305787.2 and XP_009304062.1), fruit fly PBRM1 ( M_143031.2 and

P_651288.1), and worm PBRM1 ( M_001025837.3 and P_001021008.1

and. M_001025838.2 and P_001021009.1).

Representative sequences of PBRM1 orthologs are presented below in Table 1. Anti- PBRM1 antibodies suitable for detecting PBRM1 protein are well-known in the art and include, for example, ABE70 (rabbit polyclonal antibody, EMD Millipore, Billerica, MA), TA345237 and TA345238 (rabbit polyclonal antibodies, OriGene Technologies, Rockville, MD), BP2-30673 (mouse monoclonal) and other polyclonal antibodes (Novus

Biologicals, Littleton, CO), ab 196022 (rabiit mAb, AbCam, Cambridge, MA),

PAH437Hu01 and PAH437Hu02 (rabbit polyclonal antibodies, Cloud-Clone Corp., Houston, TX), GTX100781 (GeneTex, Irvine, CA), 25-498 (ProSci, Poway, CA), sc- 367222 (Santa Cruz Biotechnology, Dallas, TX), etc. In addition, reagents are well-known for detecting PBRMl expression (see, for example, PBRMl Hu-Cy3 or Hu-Cy5

SmartFlare™ RNA Detection Probe (EMD Millipore). Moreover, mutilple siRNA, shRNA, CRISPR constructs for reducing PBRMl expression can be found in the commercial product lists of the above-referenced

companies. Ribavirin and PFI 3 are known PBRMl inhibitors. It is to be noted that the term can further be used to refer to any combination of features described herein regarding PBRMl molecules. For example, any combination of sequence composition, percentage identify, sequence length, domain structure, functional activity, etc. can be used to describe an PBRMl molecule of the present invention.

The term "PBRMl loss of function mutation" refers to any mutation in a PBRMl- related nucleic acid or protein that results in reduced or eliminated PBRMl protein amounts and/or function. For example, nucleic acid mutations include single-base substitutions, multi-base substitutions, insertion mutations, deletion mutations, frameshift mutations, missesnse mutations, nonsense mutations, splice-site mutations, epigenetic modifications {e.g., methylation, phosphorylation, acetylation, ubiquitylation, sumoylation, histone acetylation, histone deacetylation, and the like), and combinations thereof. In some embodiments, the mutation is a "nonsynonymous mutation," meaning that the mutation alters the amino acid sequence of PBRMl . Such mutations reduce or eliminate PBRMl protein amounts and/or function by eliminating proper coding sequences required for proper PBRMl protein translation and/or coding for PBRMl proteins that are nonfunctional or have reduced function {e.g., deletion of enzymatic and/or structural domains, reduction in protein stability, alteration of sub-cellular localization, and the like). Such mutations are well-known in the art. In addition, a representative list describing a wide variety of structural mutations correlated with the functional result of reduced or eliminated PBRMl protein amounts and/or function is described in Table 1 and the Examples.

Without being bound by theory, it is believed that nonsense, frameshift, and splice-site mutations are particularly amenable to PBRMl loss of function because they are known to be indicative of lack of PBRMl expression in cell lines harboring such mutations.

Unless otherwise specified here within, the terms "antibody" and "antibodies" broadly encompass naturally-occurring forms of antibodies {e.g. IgG, IgA, IgM, IgE) and recombinant antibodies such as single-chain antibodies, chimeric and humanized antibodies and multi-specific antibodies, as well as fragments and derivatives of all of the foregoing, which fragments and derivatives have at least an antigenic binding site. Antibody derivatives may comprise a protein or chemical moiety conjugated to an antibody.

The term "antibody" as used herein also includes an "antigen-binding portion" of an antibody (or simply "antibody portion"). The term "antigen-binding portion", as used herein, refers to one or more fragments of an antibody that retain the ability to specifically bind to an antigen (e.g., a biomarker polypeptide or fragment thereof). It has been shown that the antigen-binding function of an antibody can be performed by fragments of a full- length antibody. Examples of binding fragments encompassed within the term "antigen- binding portion" of an antibody include (i) a Fab fragment, a monovalent fragment consisting of the VL, VH, CL and CHI domains; (ii) a F(ab')2 fragment, a bivalent fragment comprising two Fab fragments linked by a disulfide bridge at the hinge region; (iii) a Fd fragment consisting of the VH and CHI domains; (iv) a Fv fragment consisting of the VL and VH domains of a single arm of an antibody, (v) a dAb fragment (Ward et al. (1989) Nature 341 :544-546), which consists of a VH domain; and (vi) an isolated complementarity determining region (CDR). Furthermore, although the two domains of the Fv fragment, VL and VH, are coded for by separate genes, they can be joined, using recombinant methods, by a synthetic linker that enables them to be made as a single protein chain in which the VL and VH regions pair to form monovalent polypeptides (known as single chain Fv (scFv); see e.g., Bird et al. (1988) Science 242:423-426; and Huston et al. (1988) Proc. Natl. Acad. Sci. USA 85:5879-5883; and Osbourn et al. 1998, Nature

Biotechnology 16: 778). Such single chain antibodies are also intended to be encompassed within the term "antigen-binding portion" of an antibody. Any VH and VL sequences of specific scFv can be linked to human immunoglobulin constant region cDNA or genomic sequences, in order to generate expression vectors encoding complete IgG polypeptides or other isotypes. VH and VL can also be used in the generation of Fab, Fv or other fragments of immunoglobulins using either protein chemistry or recombinant DNA technology. Other forms of single chain antibodies, such as diabodies are also encompassed. Diabodies are bivalent, bispecific antibodies in which VH and VL domains are expressed on a single polypeptide chain, but using a linker that is too short to allow for pairing between the two domains on the same chain, thereby forcing the domains to pair with complementary domains of another chain and creating two antigen binding sites (see e.g., Holliger, P., et al. (1993) Proc. Natl. Acad. Sci. USA 90:6444-6448; Poljak, R. J., et al. (1994) Structure 2: 1121-1123).

Still further, an antibody or antigen-binding portion thereof may be part of larger immunoadhesion polypeptides, formed by covalent or noncovalent association of the antibody or antibody portion with one or more other proteins or peptides. Examples of such immunoadhesion polypeptides include use of the streptavidin core region to make a tetrameric scFv polypeptide (Kipriyanov, S.M., et al. (1995) Human Antibodies and Hybridomas 6:93-101) and use of a cysteine residue, biomarker peptide and a C-terminal polyhistidine tag to make bivalent and biotinylated scFv polypeptides (Kipriyanov, S.M., et al. (1994) Mol. Immunol. 31 : 1047-1058). Antibody portions, such as Fab and F(ab')2 fragments, can be prepared from whole antibodies using conventional techniques, such as papain or pepsin digestion, respectively, of whole antibodies. Moreover, antibodies, antibody portions and immunoadhesion polypeptides can be obtained using standard recombinant DNA techniques, as described herein.

Antibodies may be polyclonal or monoclonal; xenogeneic, allogeneic, or syngeneic; or modified forms thereof (e.g. humanized, chimeric, etc.). Antibodies may also be fully human. Preferably, antibodies of the present invention bind specifically or substantially specifically to a biomarker polypeptide or fragment thereof. The terms "monoclonal antibodies" and "monoclonal antibody composition", as used herein, refer to a population of antibody polypeptides that contain only one species of an antigen binding site capable of immunoreacting with a particular epitope of an antigen, whereas the term "polyclonal antibodies" and "polyclonal antibody composition" refer to a population of antibody polypeptides that contain multiple species of antigen binding sites capable of interacting with a particular antigen. A monoclonal antibody composition typically displays a single binding affinity for a particular antigen with which it immunoreacts.

Antibodies may also be "humanized", which is intended to include antibodies made by a non-human cell having variable and constant regions which have been altered to more closely resemble antibodies that would be made by a human cell. For example, by altering the non-human antibody amino acid sequence to incorporate amino acids found in human germline immunoglobulin sequences. The humanized antibodies of the present invention may include amino acid residues not encoded by human germline immunoglobulin sequences (e.g., mutations introduced by random or site-specific mutagenesis in vitro or by somatic mutation in vivo), for example in the CDRs. The term "humanized antibody", as used herein, also includes antibodies in which CDR sequences derived from the germline of another mammalian species, such as a mouse, have been grafted onto human framework sequences.

The term "assigned score" refers to the numerical value designated for each of the biomarkers after being measured in a patient sample. The assigned score correlates to the absence, presence or inferred amount of the biomarker in the sample. The assigned score can be generated manually (e.g., by visual inspection) or with the aid of instrumentation for image acquisition and analysis. In certain embodiments, the assigned score is determined by a qualitative assessment, for example, detection of a fluorescent readout on a graded scale, or quantitative assessment. In one embodiment, an "aggregate score," which refers to the combination of assigned scores from a plurality of measured biomarkers, is determined. In one embodiment the aggregate score is a summation of assigned scores. In another embodiment, combination of assigned scores involves performing mathematical operations on the assigned scores before combining them into an aggregate score. In certain, embodiments, the aggregate score is also referred to herein as the "predictive score."

The term "biomarker" refers to a measurable entity of the present invention that has been determined to be predictive of immune checkpoint therapy effects on a cancer.

Biomarkers can include, without limitation, nucleic acids and proteins, including those shown in Table 1, the Examples, and the Figures.

A "blocking" antibody or an antibody "antagonist" is one which inhibits or reduces at least one biological activity of the antigen(s) it binds. In certain embodiments, the blocking antibodies or antagonist antibodies or fragments thereof described herein substantially or completely inhibit a given biological activity of the antigen(s).

The term "body fluid" refers to fluids that are excreted or secreted from the body as well as fluids that are normally not (e.g. amniotic fluid, aqueous humor, bile, blood and blood plasma, cerebrospinal fluid, cerumen and earwax, cowper's fluid or pre-ejaculatory fluid, chyle, chyme, stool, female ejaculate, interstitial fluid, intracellular fluid, lymph, menses, breast milk, mucus, pleural fluid, pus, saliva, sebum, semen, serum, sweat, synovial fluid, tears, urine, vaginal lubrication, vitreous humor, vomit).

The terms "cancer" or "tumor" or "hyperproliferative" refer to the presence of cells possessing characteristics typical of cancer-causing cells, such as uncontrolled proliferation, immortality, metastatic potential, rapid growth and proliferation rate, and certain

characteristic morphological features, as well as evading immune destruction (Hanahan and Weinberg (2000) 100:57-70; Hanahan and Weinberg (2011) Cell 144:646-674). In some embodiments, such cells exhibit such characteristics in part or in full due to the expression and activity of immune checkpoint proteins, such as PD-1, PD-Ll, and/or CTLA-4. Cancer cells are often in the form of a tumor, but such cells may exist alone within an animal, or may be a non-turn origenic cancer cell, such as a leukemia cell. As used herein, the term "cancer" includes premalignant as well as malignant cancers. Cancers include, but are not limited to, B cell cancer, e.g., multiple myeloma, Waldenstrom's macroglobulinemia, the heavy chain diseases, such as, for example, alpha chain disease, gamma chain disease, and mu chain disease, benign monoclonal gammopathy, and immunocytic amyloidosis, melanomas, breast cancer, lung cancer, bronchus cancer, colorectal cancer, prostate cancer, pancreatic cancer, stomach cancer, ovarian cancer, urinary bladder cancer, brain or central nervous system cancer, peripheral nervous system cancer, esophageal cancer, cervical cancer, uterine or endometrial cancer, cancer of the oral cavity or pharynx, liver cancer, kidney cancer, testicular cancer, biliary tract cancer, small bowel or appendix cancer, salivary gland cancer, thyroid gland cancer, adrenal gland cancer, osteosarcoma, chondrosarcoma, cancer of hematologic tissues, and the like. Other non-limiting examples of types of cancers applicable to the methods encompassed by the present invention include human sarcomas and carcinomas, e.g., fibrosarcoma, myxosarcoma, liposarcoma, chondrosarcoma, osteogenic sarcoma, chordoma, angiosarcoma, endotheliosarcoma, lymphangiosarcoma, lymphangioendotheliosarcoma, synovioma, mesothelioma, Ewing's tumor, leiomyosarcoma, rhabdomyosarcoma, colon carcinoma, colorectal cancer, pancreatic cancer, breast cancer, ovarian cancer, prostate cancer, squamous cell carcinoma, basal cell carcinoma, adenocarcinoma, sweat gland carcinoma, sebaceous gland carcinoma, papillary carcinoma, papillary adenocarcinomas, cystadenocarcinoma, medullary carcinoma, bronchogenic carcinoma, renal cell carcinoma, hepatoma, bile duct carcinoma, liver cancer, choriocarcinoma, seminoma, embryonal carcinoma, Wilms' tumor, cervical cancer, bone cancer, brain tumor, testicular cancer, lung carcinoma, small cell lung carcinoma, bladder carcinoma, epithelial carcinoma, glioma, astrocytoma,

medulloblastoma, craniopharyngioma, ependymoma, pinealoma, hemangioblastoma, acoustic neuroma, oligodendroglioma, meningioma, melanoma, neuroblastoma, retinoblastoma; leukemias, e.g., acute lymphocytic leukemia and acute myelocytic leukemia (myeloblastic, promyelocytic, myelomonocytic, monocytic and erythroleukemia); chronic leukemia (chronic myelocytic (granulocytic) leukemia and chronic lymphocytic leukemia); and polycythemia vera, lymphoma (Hodgkin's disease and non-Hodgkin's disease), multiple myeloma, Waldenstrom's macroglobulinemia, and heavy chain disease. In some embodiments, cancers are epithlelial in nature and include but are not limited to, bladder cancer, breast cancer, cervical cancer, colon cancer, gynecologic cancers, renal cancer, laryngeal cancer, lung cancer, oral cancer, head and neck cancer, ovarian cancer, pancreatic cancer, prostate cancer, or skin cancer. In other embodiments, the cancer is breast cancer, prostate cancer, lung cancer, or colon cancer. In still other embodiments, the epithelial cancer is non-small-cell lung cancer, nonpapillary renal cell carcinoma, cervical carcinoma, ovarian carcinoma (e.g., serous ovarian carcinoma), or breast carcinoma. The epithelial cancers may be characterized in various other ways including, but not limited to, serous, endometrioid, mucinous, clear cell, Brenner, or undifferentiated.

In certain embodiments, the cancer encompasses renal cell carcinoma (RCC). The term "renal cell carcinoma" generally refers to a type of kidney cancer that starts in the lining of the proximal convoluted tubule, a part of the very small tubes in the kidney that transport waste molecules from the blood to the urine. RCC is the most common type of kidney cancer in adults, responsible for approximately 90-95% of cases. Renal cell carcinoma is the most common type of kidney cancer in adults. It occurs most often in men 50 to 70 years old. The different types of RCC are generally distinguished by the way that cancer cells appear when viewed under a microscope, such as clear cell RCC (ccRCC), papillary RCC, chromophobe RCC, oncocytoma RCC, collecting duct RCC, and other unclassified RCC. In clear cell RCC or conventional RCC, the cells have a clear or pale appearance. CCRCC classically has apical nuclei, i.e. the nucleus is adjacent to the luminal aspect (Bing and Tomaszewski (2011) Case Rep Transplant. 2011 :387645). In most glandular structures the nuclei are usually basally located, i.e. in the cytoplasm adjacent to the basement membrane. They typically stain with CK7 and do not stain with TFE3 and AMACR (Rohan et al. (2011) Mod Pathol. 24: 1207-1220). Around 70 to 80 percent of individuals with renal cell cancer have clear cell RCC. The growth of these cells can be either slow or fast. Metastatic renal cell carcinoma (mRCC) is the spread of the primary renal cell carcinoma from the kidney to other organs. About 25-30% of people have this metastatic spread by the time they are diagnosed with renal cell carcinoma. This high proportion is explained by the fact that clinical signs are generally mild until the disease progresses to a more severe state. The most common sites for metastasis are the lymph nodes, lung, bones, liver and brain. mRCC has a poor prognosis compared to other cancers, though average survival times have increased in the last few years due to treatment advances. Average survival time in 2008 for the metastatic form of the disease was under a year and by 2013 this improved to an average of 22 months. Despite this improvement, the 5-year survival rate for mRCC remains under 10%. About 20-25%) of suffers remain unresponsive to all treatments and in these cases, the disease has a rapid progression. The known risk factors of kidney cancer include, e.g., smoking, obesity, dialysis treatment, family history of the disease, high blood pressure, horseshoe kidney, long-term use of certain medicines, such as pain pills or water pills (diuretics), polycystic kidney disease, von Hippel-Lindau disease (a hereditary disease that affects blood vessels in the brain, eyes, and other body parts), etc. Symptoms of RCC may include any of the following: abdominal pain and swelling, back pain, blood in the urine, swelling of the veins around a testicle (varicocele), flank pain, weight loss, excessive hair growth in females, pale skin, vision problems, etc. The initial symptoms of RCC often include: blood in the urine (occurring in 40% of affected persons at the time they first seek medical attention), flank pain (40%), a mass in the abdomen or flank (25%), weight loss (33%), fever (20%), high blood pressure (20%)), night sweats and generally feeling unwell. When RCC metastasises, it most commonly spreads to the lymph nodes, lungs, liver, adrenal glands, brain or bones. RCC is also associated with a number of paraneoplastic syndromes (PNS) which are conditions caused by either the hormones produced by the tumour or by the body's attack on the tumour and are present in about 20% of those with RCC. These paraneoplastic syndromes most commonly affect tissues which have not been invaded by the cancer. The most common PNSs seen in people with RCC are: high blood calcium levels, polycythaemia (the opposite of anaemia, due to an overproduction of erythropoietin), thrombocytosis (too many platelets in the blood, leading to an increased tendency for blood clotting and bleeds) and secondary amyloidosis. For exam and diagnosis, a physical exam may reveal mass or swelling of the abdomen and/or a varicocele in the male scrotum. Diagnostic tests include, e.g., abdominal CT scan, blood chemistry, complete blood count (CBC), intravenous pyelogram (IVP), liver function tests, renal arteriography, ultrasound of the abdomen and kidney, and urine tests. Tests for detecting spread RCC may include abdominal CT scan, adominal MRI, bone scan, chest x-ray or CT scan, and PET scan. Availabe treatment for RCC may include surgery to remove of all or part of the kidney (nephrectomy). This may include removing the bladder, surrounding tissues, or lymph nodes. Chemotherapy or radiation therapy is generally not effective for treating kidney cancer. Current

immunotherapies include the immune system medicines interleukin-2 (IL-2) and

nivolumab, developed after observing that in some cases there was spontaneous regression (Davar et al. (2013) "Immunotherapy for Renal Cell Carcinoma". Renal Cell Carcinoma Clinical Management. Humana, pp. 279-302). Other targeted therapies include anti- angiogenesis therapies {e.g., bevacizumab (Avastin ^® )), tyrosine kinase inhibitors (TKIs) {e.g., cabozantinib (Cabometyx ^™ ), pazopanib (Votrient ^® ), sorafenib (Nexavar ^® ), axitinib (INLYTA ^® ) and sunitinib (Sutent ^® )), mTOR inhibitors (e.g., Everolimus (Afinitor ^® ) and temsirolimus (Torisel ^® )), and other inhibitors to growth factors that have been shown to promote the growth and spread of tumours (e.g., lenvatinib (LENVIMA ^® ), also see Santoni et al. (2013) Expert Review of Anticancer Therapy. 13 :697-709; Stroup (2013)

"Neoadjuvant Targeted Therapy and Consolidative Surgery" Renal Cell Carcinoma Clinical Management. Humana, pp. 219-230).

The term "coding region" refers to regions of a nucleotide sequence comprising codons which are translated into amino acid residues, whereas the term "noncoding region" refers to regions of a nucleotide sequence that are not translated into amino acids {e.g., 5' and 3' untranslated regions).

The term "complementary" refers to the broad concept of sequence

complementarity between regions of two nucleic acid strands or between two regions of the same nucleic acid strand. It is known that an adenine residue of a first nucleic acid region is capable of forming specific hydrogen bonds ("base pairing") with a residue of a second nucleic acid region which is antiparallel to the first region if the residue is thymine or uracil. Similarly, it is known that a cytosine residue of a first nucleic acid strand is capable of base pairing with a residue of a second nucleic acid strand which is antiparallel to the first strand if the residue is guanine. A first region of a nucleic acid is complementary to a second region of the same or a different nucleic acid if, when the two regions are arranged in an antiparallel fashion, at least one nucleotide residue of the first region is capable of base pairing with a residue of the second region. Preferably, the first region comprises a first portion and the second region comprises a second portion, whereby, when the first and second portions are arranged in an antiparallel fashion, at least about 50%, and preferably at least about 75%, at least about 90%, or at least about 95% of the nucleotide residues of the first portion are capable of base pairing with nucleotide residues in the second portion.

More preferably, all nucleotide residues of the first portion are capable of base pairing with nucleotide residues in the second portion.

The terms "conjoint therapy" and "combination therapy," as used herein, refer to the administration of two or more therapeutic substances, e.g., combinations of anti-immune checkpoint therapies, multiple inhibitors of an immune checkpoint of interest, combinations of immune checkpoint therapy with an inhibitor of PBRMl, and combinations thereof. The different agents comprising the combination therapy may be administered concomitant with, prior to, or following the administration of one or more therapeutic agents. The term "control" refers to any reference standard suitable to provide a comparison to the expression products in the test sample. In one embodiment, the control comprises obtaining a "control sample" from which expression product levels are detected and compared to the expression product levels from the test sample. Such a control sample may comprise any suitable sample, including but not limited to a sample from a control cancer patient (can be stored sample or previous sample measurement) with a known outcome; normal tissue or cells isolated from a subject, such as a normal patient or the cancer patient, cultured primary cells/tissues isolated from a subject such as a normal subject or the cancer patient, adjacent normal cells/tissues obtained from the same organ or body location of the cancer patient, a tissue or cell sample isolated from a normal subject, or a primary cells/tissues obtained from a depository. In another preferred embodiment, the control may comprise a reference standard expression product level from any suitable source, including but not limited to housekeeping genes, an expression product level range from normal tissue (or other previously analyzed control sample), a previously determined expression product level range within a test sample from a group of patients, or a set of patients with a certain outcome (for example, survival for one, two, three, four years, etc.) or receiving a certain treatment (for example, standard of care cancer therapy). It will be understood by those of skill in the art that such control samples and reference standard expression product levels can be used in combination as controls in the methods of the present invention. In one embodiment, the control may comprise normal or non-cancerous cell/tissue sample. In another preferred embodiment, the control may comprise an expression level for a set of patients, such as a set of cancer patients, or for a set of cancer patients receiving a certain treatment, or for a set of patients with one outcome versus another outcome. In the former case, the specific expression product level of each patient can be assigned to a percentile level of expression, or expressed as either higher or lower than the mean or average of the reference standard expression level. In another preferred embodiment, the control may comprise normal cells, cells from patients treated with combination chemotherapy, and cells from patients having benign cancer. In another embodiment, the control may also comprise a measured value for example, average level of expression of a particular gene in a population compared to the level of expression of a housekeeping gene in the same population. Such a population may comprise normal subjects, cancer patients who have not undergone any treatment (i.e., treatment naive), cancer patients undergoing standard of care therapy, or patients having benign cancer. In another preferred embodiment, the control comprises a ratio transformation of expression product levels, including but not limited to determining a ratio of expression product levels of two genes in the test sample and comparing it to any suitable ratio of the same two genes in a reference standard;

determining expression product levels of the two or more genes in the test sample and determining a difference in expression product levels in any suitable control; and determining expression product levels of the two or more genes in the test sample, normalizing their expression to expression of housekeeping genes in the test sample, and comparing to any suitable control. In particularly preferred embodiments, the control comprises a control sample which is of the same lineage and/or type as the test sample. In another embodiment, the control may comprise expression product levels grouped as percentiles within or based on a set of patient samples, such as all patients with cancer. In one embodiment a control expression product level is established wherein higher or lower levels of expression product relative to, for instance, a particular percentile, are used as the basis for predicting outcome. In another preferred embodiment, a control expression product level is established using expression product levels from cancer control patients with a known outcome, and the expression product levels from the test sample are compared to the control expression product level as the basis for predicting outcome. As demonstrated by the data below, the methods of the present invention are not limited to use of a specific cut-point in comparing the level of expression product in the test sample to the control.

The "copy number" of a biomarker nucleic acid refers to the number of DNA sequences in a cell (e.g., germline and/or somatic) encoding a particular gene product. Generally, for a given gene, a mammal has two copies of each gene. The copy number can be increased, however, by gene amplification or duplication, or reduced by deletion. For example, germline copy number changes include changes at one or more genomic loci, wherein said one or more genomic loci are not accounted for by the number of copies in the normal complement of germline copies in a control (e.g., the normal copy number in germline DNA for the same species as that from which the specific germline DNA and corresponding copy number were determined). Somatic copy number changes include changes at one or more genomic loci, wherein said one or more genomic loci are not accounted for by the number of copies in germline DNA of a control (e.g., copy number in germline DNA for the same subject as that from which the somatic DNA and corresponding copy number were determined). The "normal" copy number (e.g., germline and/or somatic) of a biomarker nucleic acid or "normal" level of expression of a biomarker nucleic acid or protein is the activity /level of expression or copy number in a biological sample, e.g., a sample containing tissue, whole blood, serum, plasma, buccal scrape, saliva, cerebrospinal fluid, urine, stool, and bone marrow, from a subject, e.g., a human, not afflicted with cancer, or from a corresponding non-cancerous tissue in the same subject who has cancer.

As used herein, the term "costimulate" with reference to activated immune cells includes the ability of a costimulatory molecule to provide a second, non-activating receptor mediated signal (a "costimulatory signal") that induces proliferation or effector function. For example, a costimulatory signal can result in cytokine secretion, e.g., in a T cell that has received a T cell-receptor-mediated signal. Immune cells that have received a cell-receptor mediated signal, e.g., via an activating receptor are referred to herein as "activated immune cells."

The term "determining a suitable treatment regimen for the subject" is taken to mean the determination of a treatment regimen (i.e., a single therapy or a combination of different therapies that are used for the prevention and/or treatment of the cancer in the subject) for a subject that is started, modified and/or ended based or essentially based or at least partially based on the results of the analysis according to the present invention. One example is determining whether to provide targeted therapy against a cancer to provide immunotherapy that generally increases immune responses against the cancer (e.g. , immune checkpoint therapy). Another example is starting an adjuvant therapy after surgery whose purpose is to decrease the risk of recurrence, another would be to modify the dosage of a particular chemotherapy. The determination can, in addition to the results of the analysis according to the present invention, be based on personal characteristics of the subject to be treated. In most cases, the actual determination of the suitable treatment regimen for the subject will be performed by the attending physician or doctor.

The term "diagnosing cancer" includes the use of the methods, systems, and code of the present invention to determine the presence or absence of a cancer or subtype thereof in an individual. The term also includes methods, systems, and code for assessing the level of disease activity in an individual.

A molecule is "fixed" or "affixed" to a substrate if it is covalently or non-covalently associated with the substrate such that the substrate can be rinsed with a fluid (e.g. standard saline citrate, pH 7.4) without a substantial fraction of the molecule dissociating from the substrate.

The term "expression signature" or "signature" refers to a group of two or more coordinately expressed biomarkers. For example, the genes, proteins, metabolites, and the like making up this signature may be expressed in a specific cell lineage, stage of differentiation, or during a particular biological response. The biomarkers can reflect biological aspects of the tumors in which they are expressed, such as the cell of origin of the cancer, the nature of the non-malignant cells in the biopsy, and the oncogenic mechanisms responsible for the cancer. Expression data and gene expression levels can be stored on computer readable media, e.g., the computer readable medium used in

conjunction with a microarray or chip reading device. Such expression data can be manipulated to generate expression signatures.

"Homologous" as used herein, refers to nucleotide sequence similarity between two regions of the same nucleic acid strand or between regions of two different nucleic acid strands. When a nucleotide residue position in both regions is occupied by the same nucleotide residue, then the regions are homologous at that position. A first region is homologous to a second region if at least one nucleotide residue position of each region is occupied by the same residue. Homology between two regions is expressed in terms of the proportion of nucleotide residue positions of the two regions that are occupied by the same nucleotide residue. By way of example, a region having the nucleotide sequence 5'- ATTGCC-3' and a region having the nucleotide sequence 5'-TATGGC-3' share 50% homology. Preferably, the first region comprises a first portion and the second region comprises a second portion, whereby, at least about 50%, and preferably at least about 75%, at least about 90%, or at least about 95% of the nucleotide residue positions of each of the portions are occupied by the same nucleotide residue. More preferably, all nucleotide residue positions of each of the portions are occupied by the same nucleotide residue.

The term "immune cell" refers to cells that play a role in the immune response. Immune cells are of hematopoietic origin, and include lymphocytes, such as B cells and T cells; natural killer cells; myeloid cells, such as monocytes, macrophages, eosinophils, mast cells, basophils, and granulocytes.

The term "immune checkpoint" refers to a group of molecules on the cell surface of CD4+ and/or CD8+ T cells that fine-tune immune responses by down-modulating or inhibiting an anti-tumor immune response. Immune checkpoint proteins are well known in the art and include, without limitation, CTLA-4, PD-1, VISTA, B7-H2, B7-H3, PD-Ll, B7- H4, B7-H6, 2B4, ICOS, HVEM, PD-L2, CD160, gp49B, PIR-B, KIR family receptors, TIM-1, TIM-3, TIM-4, LAG-3, BTLA, SIRPalpha (CD47), CD48, 2B4 (CD244), B7.1, B7.2, ILT-2, ILT-4, TIGIT, and A2aR (see, for example, WO 2012/177624). The term further encompasses biologically active protein fragment, as well as nucleic acids encoding full-length immune checkpoint proteins and biologically active protein fragments thereof. In some embodiment, the term further encompasses any fragment according to homology descriptions provided herein.

"Immune checkpoint therapy" refers to the use of agents that inhibit immune checkpoint nucleic acids and/or proteins. Inhibition of one or more immune checkpoints can block or otherwise neutralize inhibitory signaling to thereby upregulate an immune response in order to more efficaciously treat cancer. Exemplary agents useful for inhibiting immune checkpoints include antibodies, small molecules, peptides, peptidomimetics, natural ligands, and derivatives of natural ligands, that can either bind and/or inactivate or inhibit immune checkpoint proteins, or fragments thereof; as well as RNA interference, antisense, nucleic acid aptamers, etc. that can downregulate the expression and/or activity of immune checkpoint nucleic acids, or fragments thereof. Exemplary agents for upregulating an immune response include antibodies against one or more immune checkpoint proteins block the interaction between the proteins and its natural receptor(s); a non-activating form of one or more immune checkpoint proteins (e.g. , a dominant negative polypeptide); small molecules or peptides that block the interaction between one or more immune checkpoint proteins and its natural receptor(s); fusion proteins (e.g. the

extracellular portion of an immune checkpoint inhibition protein fused to the Fc portion of an antibody or immunoglobulin) that bind to its natural receptor(s); nucleic acid molecules that block immune checkpoint nucleic acid transcription or translation; and the like. Such agents can directly block the interaction between the one or more immune checkpoints and its natural receptor(s) (e.g., antibodies) to prevent inhibitory signaling and upregulate an immune response. Alternatively, agents can indirectly block the interaction between one or more immune checkpoint proteins and its natural receptor(s) to prevent inhibitory signaling and upregulate an immune response. For example, a soluble version of an immune checkpoint protein ligand such as a stabilized extracellular domain can binding to its receptor to indirectly reduce the effective concentration of the receptor to bind to an appropriate ligand. In one embodiment, anti-PD-1 antibodies, anti -PD-Ll antibodies, and anti-CTLA-4 antibodies, either alone or in combination, are used to inhibit immune checkpoints.

"Ipilimumab" is a representative example of an immune checkpoint therapy.

Ipilimumab (previously MDX-010; Medarex Inc., marketed by Bristol-Myers Squibb as YERVOY™) is a fully human anti-human CTLA-4 monoclonal antibody that blocks the binding of CTLA-4 to CD80 and CD86 expressed on antigen presenting cells, thereby, blocking the negative down-regulation of the immune responses elicited by the interaction of these molecules (see, for example, WO 2013/169971, U.S. Pat. Publ. 2002/0086014, and U.S. Pat. Publ. 2003/0086930.

The term "immune response" includes T cell mediated and/or B cell mediated immune responses. Immune responses can also include B- and T-cell independent and rely on macrophages and NK cells (along with other cell types) instead (innate immunity). Exemplary immune responses include T cell responses, e.g., cytokine production and cellular cytotoxicity. In addition, the term immune response includes immune responses that are indirectly effected by T cell activation, e.g., antibody production (humoral responses) and activation of cytokine responsive cells, e.g., macrophages.

The term "immunotherapeutic agent" can include any molecule, peptide, antibody or other agent which can stimulate a host immune system to generate an immune response to a tumor or cancer in the subject. Various immunotherapeutic agents are useful in the compositions and methods described herein.

The term "inhibit" includes the decrease, limitation, or blockage, of, for example a particular action, function, or interaction. In some embodiments, cancer is "inhibited" if at least one symptom of the cancer is alleviated, terminated, slowed, or prevented. As used herein, cancer is also "inhibited" if recurrence or metastasis of the cancer is reduced, slowed, delayed, or prevented.

The term "interaction", when referring to an interaction between two molecules, refers to the physical contact (e.g., binding) of the molecules with one another. Generally, such an interaction results in an activity (which produces a biological effect) of one or both of said molecules.

An "isolated protein" refers to a protein that is substantially free of other proteins, cellular material, separation medium, and culture medium when isolated from cells or produced by recombinant DNA techniques, or chemical precursors or other chemicals when chemically synthesized. An "isolated" or "purified" protein or biologically active portion thereof is substantially free of cellular material or other contaminating proteins from the cell or tissue source from which the antibody, polypeptide, peptide or fusion protein is derived, or substantially free from chemical precursors or other chemicals when chemically synthesized. The language "substantially free of cellular material" includes preparations of a biomarker polypeptide or fragment thereof, in which the protein is separated from cellular components of the cells from which it is isolated or recombinantly produced. In one embodiment, the language "substantially free of cellular material" includes preparations of a biomarker protein or fragment thereof, having less than about 30% (by dry weight) of non-biomarker protein (also referred to herein as a "contaminating protein"), more preferably less than about 20% of non-biomarker protein, still more preferably less than about 10%) of non-biomarker protein, and most preferably less than about 5% non- biomarker protein. When antibody, polypeptide, peptide or fusion protein or fragment thereof, e.g., a biologically active fragment thereof, is recombinantly produced, it is also preferably substantially free of culture medium, i.e., culture medium represents less than about 20%, more preferably less than about 10%>, and most preferably less than about 5% of the volume of the protein preparation.

A "kit" is any manufacture (e.g. a package or container) comprising at least one reagent, e.g. a probe or small molecule, for specifically detecting and/or affecting the expression of a marker of the present invention. The kit may be promoted, distributed, or sold as a unit for performing the methods of the present invention. The kit may comprise one or more reagents necessary to express a composition useful in the methods of the present invention. In certain embodiments, the kit may further comprise a reference standard, e.g., a nucleic acid encoding a protein that does not affect or regulate signaling pathways controlling cell growth, division, migration, survival or apoptosis. One skilled in the art can envision many such control proteins, including, but not limited to, common molecular tags (e.g., green fluorescent protein and beta-galactosidase), proteins not classified in any of pathway encompassing cell growth, division, migration, survival or apoptosis by GeneOntology reference, or ubiquitous housekeeping proteins. Reagents in the kit may be provided in individual containers or as mixtures of two or more reagents in a single container. In addition, instructional materials which describe the use of the compositions within the kit can be included.

The term "neoadjuvant therapy" refers to a treatment given before the primary treatment. Examples of neoadjuvant therapy can include chemotherapy, radiation therapy, and hormone therapy. For example, in treating breast cancer, neoadjuvant therapy can allows patients with large breast cancer to undergo breast-conserving surgery.

The "normal" level of expression of a biomarker is the level of expression of the biomarker in cells of a subject, e.g., a human patient, not afflicted with a cancer. An "over- expression" or "significantly higher level of expression" of a biomarker refers to an expression level in a test sample that is greater than the standard error of the assay employed to assess expression, and is preferably at least 10%, and more preferably 1.2, 1.3,

1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 2.9, 3, 3.5, 4, 4.5, 5,

5.5, 6, 6.5, 7, 7.5, 8, 8.5, 9, 9.5, 10, 10.5, 1 1, 12, 13, 14, 15, 16, 17, 18, 19, 20 times or more higher than the expression activity or level of the biomarker in a control sample (e.g., sample from a healthy subject not having the biomarker associated disease) and preferably, the average expression level of the biomarker in several control samples. A "significantly lower level of expression" of a biomarker refers to an expression level in a test sample that is at least 10%, and more preferably 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 2.9, 3, 3.5, 4, 4.5, 5, 5.5, 6, 6.5, 7, 7.5, 8, 8.5, 9, 9.5, 10, 10.5, 1 1, 12, 13, 14, 15, 16, 17, 18, 19, 20 times or more lower than the expression level of the biomarker in a control sample (e.g., sample from a healthy subject not having the biomarker associated disease) and preferably, the average expression level of the biomarker in several control samples.

An "over-expression" or "significantly higher level of expression" of a biomarker refers to an expression level in a test sample that is greater than the standard error of the assay employed to assess expression, and is preferably at least 10%, and more preferably 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 2.9, 3, 3.5, 4, 4.5, 5, 5.5, 6, 6.5, 7, 7.5, 8, 8.5, 9, 9.5, 10, 10.5, 1 1, 12, 13, 14, 15, 16, 17, 18, 19, 20 times or more higher than the expression activity or level of the biomarker in a control sample (e.g., sample from a healthy subject not having the biomarker associated disease) and preferably, the average expression level of the biomarker in several control samples. A "significantly lower level of expression" of a biomarker refers to an expression level in a test sample that is at least 10%, and more preferably 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 2.9, 3, 3.5, 4, 4.5, 5, 5.5, 6, 6.5, 7, 7.5, 8, 8.5, 9, 9.5, 10, 10.5, 1 1, 12, 13, 14, 15, 16, 17, 18, 19, 20 times or more lower than the expression level of the biomarker in a control sample (e.g., sample from a healthy subject not having the biomarker associated disease) and preferably, the average expression level of the biomarker in several control samples.

The term "pre-determined" biomarker amount and/or activity measurement(s) may be a biomarker amount and/or activity measurement(s) used to, by way of example only, evaluate a subject that may be selected for a particular treatment, evaluate a response to a treatment such as an anti-immune checkpoint inhibitor therapy, and/or evaluate the disease state. A pre-determined biomarker amount and/or activity measurement(s) may be determined in populations of patients with or without cancer. The pre-determined biomarker amount and/or activity measurement(s) can be a single number, equally applicable to every patient, or the pre-determined biomarker amount and/or activity measurement(s) can vary according to specific subpopulations of patients. Age, weight, height, and other factors of a subject may affect the pre-determined biomarker amount and/or activity measurement(s) of the individual. Furthermore, the pre-determined biomarker amount and/or activity can be determined for each subject individually. In one embodiment, the amounts determined and/or compared in a method described herein are based on absolute measurements. In another embodiment, the amounts determined and/or compared in a method described herein are based on relative measurements, such as ratios (e.g., serum biomarker normalized to the expression of a housekeeping or otherwise generally constant biomarker). The pre-determined biomarker amount and/or activity measurement s) can be any suitable standard. For example, the pre-determined biomarker amount and/or activity measurement(s) can be obtained from the same or a different human for whom a patient selection is being assessed. In one embodiment, the pre-determined biomarker amount and/or activity measurement(s) can be obtained from a previous assessment of the same patient. In such a manner, the progress of the selection of the patient can be monitored over time. In addition, the control can be obtained from an assessment of another human or multiple humans, e.g., selected groups of humans, if the subject is a human. In such a manner, the extent of the selection of the human for whom selection is being assessed can be compared to suitable other humans, e.g., other humans who are in a similar situation to the human of interest, such as those suffering from similar or the same condition(s) and/or of the same ethnic group.

The term "predictive" includes the use of a biomarker nucleic acid and/or protein status, e.g., over- or under- activity, emergence, expression, growth, remission, recurrence or resistance of tumors before, during or after therapy, for determining the likelihood of response of a cancer to anti -immune checkpoint treatment (e.g., therapeutic antibodies against CTLA-4, PD-1, PD-Ll, and the like). Such predictive use of the biomarker may be confirmed by, e.g., (1) increased or decreased copy number (e.g., by FISH, FISH plus SKY, single-molecule sequencing, e.g., as described in the art at least at J. Biotechnol., 86:289- 301, or qPCR), overexpression or underexpression of a biomarker nucleic acid (e.g., by ISH, Northern Blot, or qPCR), increased or decreased biomarker protein (e.g., by IHC), or increased or decreased activity, e.g., in more than about 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, 100%, or more of assayed human cancers types or cancer samples; (2) its absolute or relatively modulated presence or absence in a biological sample, e.g., a sample containing tissue, whole blood, serum, plasma, buccal scrape, saliva, cerebrospinal fluid, urine, stool, or bone marrow, from a subject, e.g. a human, afflicted with cancer; (3) its absolute or relatively modulated presence or absence in clinical subset of patients with cancer (e.g., those responding to a particular immune checkpoint therapy or those developing resistance thereto).

The term "pre-malignant lesions" as described herein refers to a lesion that, while not cancerous, has potential for becoming cancerous. It also includes the term "pre- malignant disorders" or "potentially malignant disorders." In particular this refers to a benign, morphologically and/or histologically altered tissue that has a greater than normal risk of malignant transformation, and a disease or a patient's habit that does not necessarily alter the clinical appearance of local tissue but is associated with a greater than normal risk of precancerous lesion or cancer development in that tissue (leukoplakia, erythroplakia, erytroleukoplakia lichen planus (lichenoid reaction) and any lesion or an area which histological examination showed atypia of cells or dysplasia.

The terms "prevent," "preventing," "prevention," "prophylactic treatment," and the like refer to reducing the probability of developing a disease, disorder, or condition in a subject, who does not have, but is at risk of or susceptible to developing a disease, disorder, or condition.

The term "probe" refers to any molecule which is capable of selectively binding to a specifically intended target molecule, for example, a nucleotide transcript or protein encoded by or corresponding to a biomarker nucleic acid. Probes can be either synthesized by one skilled in the art, or derived from appropriate biological preparations. For purposes of detection of the target molecule, probes may be specifically designed to be labeled, as described herein. Examples of molecules that can be utilized as probes include, but are not limited to, RNA, DNA, proteins, antibodies, and organic molecules.

The term "prognosis" includes a prediction of the probable course and outcome of cancer or the likelihood of recovery from the disease. In some embodiments, the use of statistical algorithms provides a prognosis of cancer in an individual. For example, the prognosis can be surgery, development of a clinical subtype of cancer (e.g., solid tumors, such as lung cancer, melanoma, and renal cell carcinoma), development of one or more clinical factors, development of intestinal cancer, or recovery from the disease.

The term "response to immune checkpoint therapy" relates to any response of the hyperproliferative disorder (e.g., cancer) to an immune checkpoint therapy, preferably to a change in tumor mass and/or volume after initiation of neoadjuvant or adjuvant

chemotherapy or as prolonged patient survival following treatment compared to patients not receiving the therapy. Hyperproliferative disorder response may be assessed, for example for efficacy or in a neoadjuvant or adjuvant situation, where the size of a tumor after systemic intervention can be compared to the initial size and dimensions as measured by CT, PET, mammogram, ultrasound or palpation. Responses may also be assessed by caliper measurement or pathological examination of the tumor after biopsy or surgical resection. Response may be recorded in a quantitative fashion like percentage change in tumor volume or in a qualitative fashion like "pathological complete response" (pCR), "clinical complete remission" (cCR), "clinical partial remission" (cPR), "clinical stable disease" (cSD), "clinical progressive disease" (cPD) or other qualitative criteria.

Assessment of hyperproliferative disorder response may be done early after the onset of neoadjuvant or adjuvant therapy, e.g., after a few hours, days, weeks or preferably after a few months. A typical endpoint for response assessment is upon termination of

neoadjuvant chemotherapy or upon surgical removal of residual tumor cells and/or the tumor bed. This is typically three months after initiation of neoadjuvant therapy. In some embodiments, clinical efficacy of the therapeutic treatments described herein may be determined by measuring the clinical benefit rate (CBR). The clinical benefit rate is measured by determining the sum of the percentage of patients who are in complete remission (CR), the number of patients who are in partial remission (PR) and the number of patients having stable disease (SD) at a time point at least 6 months out from the end of therapy. The shorthand for this formula is CBR=CR+PR+SD over 6 months. In some embodiments, the CBR for a particular cancer therapeutic regimen is at least 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, or more. Additional criteria for evaluating the response to cancer therapies are related to "survival," which includes all of the following: survival until mortality, also known as overall survival (wherein said mortality may be either irrespective of cause or tumor related); "recurrence- free survival" (wherein the term recurrence shall include both localized and distant recurrence); metastasis free survival; disease free survival (wherein the term disease shall include cancer and diseases associated therewith). The length of said survival may be calculated by reference to a defined start point (e.g., time of diagnosis or start of treatment) and end point (e.g., death, recurrence or metastasis). In addition, criteria for efficacy of treatment can be expanded to include response to chemotherapy, probability of survival, probability of metastasis within a given time period, and probability of tumor recurrence. For example, in order to determine appropriate threshold values, a particular cancer therapeutic regimen can be administered to a population of subjects and the outcome can be correlated to biomarker measurements that were determined prior to administration of any cancer therapy. The outcome measurement may be pathologic response to therapy given in the neoadjuvant setting. Alternatively, outcome measures, such as overall survival and disease-free survival can be monitored over a period of time for subjects following cancer therapy for whom biomarker measurement values are known. In certain embodiments, the doses administered are standard doses known in the art for cancer therapeutic agents. The period of time for which subjects are monitored can vary. For example, subjects may be monitored for at least 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 25, 30, 35, 40, 45, 50, 55, or 60 months. Biomarker measurement threshold values that correlate to outcome of a cancer therapy can be determined using well-known methods in the art, such as those described in the Examples section.

The term "resistance" refers to an acquired or natural resistance of a cancer sample or a mammal to a cancer therapy ( i.e., being nonresponsive to or having reduced or limited response to the therapeutic treatment), such as having a reduced response to a therapeutic treatment (cessation of tumor shrinkage and development of tumor growth while receiving a given therapy) by 25% or more, for example, 30%, 40%, 50%, 60%, 70%, 80%, or more, to 2-fold, 3-fold, 4-fold, 5-fold, 10-fold, 15-fold, 20-fold or more. The reduction in response can be measured by comparing with the same cancer sample or mammal before the resistance is acquired, or by comparing with a different cancer sample or a mammal who is known to have no resistance to the therapeutic treatment. A typical acquired resistance to chemotherapy is called "multidrug resistance." The multidrug resistance can be mediated by P-glycoprotein or can be mediated by other mechanisms, or it can occur when a mammal is infected with a multi-drug-resistant microorganism or a combination of microorganisms. The determination of resistance to a therapeutic treatment is routine in the art and within the skill of an ordinarily skilled clinician, for example, can be measured by cell proliferative assays and cell death assays as described herein as "sensitizing." In some embodiments, the term "reverses resistance" means that the use of a second agent in combination with a primary cancer therapy (e.g., chemotherapeutic or radiation therapy) is able to produce a significant decrease in tumor volume at a level of statistical significance (e.g., p<0.05) when compared to tumor volume of untreated tumor in the circumstance where the primary cancer therapy (e.g., chemotherapeutic or radiation therapy) alone is unable to produce a statistically significant decrease in tumor volume compared to tumor volume of untreated tumor. This generally applies to tumor volume measurements made at a time when the untreated tumor is growing log rhythmically.

The terms "response" or "responsiveness" refers to an anti-cancer response, e.g. in the sense of reduction of tumor size or inhibiting tumor growth. The terms can also refer to an improved prognosis, for example, as reflected by an increased time to recurrence, which is the period to first recurrence censoring for second primary cancer as a first event or death without evidence of recurrence, or an increased overall survival, which is the period from treatment to death from any cause. To respond or to have a response means there is a beneficial endpoint attained when exposed to a stimulus. Alternatively, a negative or detrimental symptom is minimized, mitigated or attenuated on exposure to a stimulus. It will be appreciated that evaluating the likelihood that a tumor or subject will exhibit a favorable response is equivalent to evaluating the likelihood that the tumor or subject will not exhibit favorable response (i.e., will exhibit a lack of response or be non-responsive).

An "RNA interfering agent" as used herein, is defined as any agent which interferes with or inhibits expression of a target biomarker gene by RNA interference (RNAi). Such RNA interfering agents include, but are not limited to, nucleic acid molecules including RNA molecules which are homologous to the target biomarker gene of the present invention, or a fragment thereof, short interfering RNA (siRNA), and small molecules which interfere with or inhibit expression of a target biomarker nucleic acid by RNA interference (RNAi). "RNA interference (RNAi)" is an evolutionally conserved process whereby the expression or introduction of RNA of a sequence that is identical or highly similar to a target biomarker nucleic acid results in the sequence specific degradation or specific post- transcriptional gene silencing (PTGS) of messenger RNA (mRNA) transcribed from that targeted gene (see Coburn, G. and Cullen, B. (2002) J. of Virology 76(18):9225), thereby inhibiting expression of the target biomarker nucleic acid. In one embodiment, the RNA is double stranded RNA (dsRNA). This process has been described in plants, invertebrates, and mammalian cells. In nature, RNAi is initiated by the dsRNA-specific endonuclease Dicer, which promotes processive cleavage of long dsRNA into double-stranded fragments termed siRNAs. siRNAs are incorporated into a protein complex that recognizes and cleaves target mRNAs. RNAi can also be initiated by introducing nucleic acid molecules, e.g., synthetic siRNAs or RNA interfering agents, to inhibit or silence the expression of target biomarker nucleic acids. As used herein, "inhibition of target biomarker nucleic acid expression" or "inhibition of marker gene expression" includes any decrease in expression or protein activity or level of the target biomarker nucleic acid or protein encoded by the target biomarker nucleic acid. The decrease may be of at least 30%, 40%, 50%, 60%, 70%, 80%), 90%), 95%) or 99% or more as compared to the expression of a target biomarker nucleic acid or the activity or level of the protein encoded by a target biomarker nucleic acid which has not been targeted by an RNA interfering agent.

The term "sample" used for detecting or determining the presence or level of at least one biomarker is typically whole blood, plasma, serum, saliva, urine, stool (e.g., feces), tears, and any other bodily fluid (e.g., as described above under the definition of "body fluids"), or a tissue sample (e.g., biopsy) such as a small intestine, colon sample, or surgical resection tissue. In certain instances, the method of the present invention further comprises obtaining the sample from the individual prior to detecting or determining the presence or level of at least one marker in the sample.

The term "sensitize" means to alter cancer cells or tumor cells in a way that allows for more effective treatment of the associated cancer with a cancer therapy (e.g., anti- immune checkpoint, chemotherapeutic, and/or radiation therapy). In some embodiments, normal cells are not affected to an extent that causes the normal cells to be unduly injured by the immune checkpoint therapy. An increased sensitivity or a reduced sensitivity to a therapeutic treatment is measured according to a known method in the art for the particular treatment and methods described herein below, including, but not limited to, cell proliferative assays (Tanigawa N, Kern D H, Kikasa Y, Morton D L, Cancer Res 1982; 42: 2159-2164), cell death assays (Weisenthal L M, Shoemaker R H, Marsden J A, Dill P L, Baker J A, Moran E M, Cancer Res 1984; 94: 161-173; Weisenthal L M, Lippman M E, Cancer Treat Rep 1985; 69: 615-632; Weisenthal L M, In: Kaspers G J L, Pieters R, Twentyman P R, Weisenthal L M, Veerman A J P, eds. Drug Resistance in Leukemia and Lymphoma. Langhorne, P A: Harwood Academic Publishers, 1993 : 415-432; Weisenthal L M, Contrib Gynecol Obstet 1994; 19: 82-90). The sensitivity or resistance may also be measured in animal by measuring the tumor size reduction over a period of time, for example, 6 month for human and 4-6 weeks for mouse. A composition or a method sensitizes response to a therapeutic treatment if the increase in treatment sensitivity or the reduction in resistance is 25% or more, for example, 30%, 40%, 50%, 60%, 70%, 80%, or more, to 2-fold, 3-fold, 4-fold, 5-fold, 10-fold, 15-fold, 20-fold or more, compared to treatment sensitivity or resistance in the absence of such composition or method. The determination of sensitivity or resistance to a therapeutic treatment is routine in the art and within the skill of an ordinarily skilled clinician. It is to be understood that any method described herein for enhancing the efficacy of a cancer therapy can be equally applied to methods for sensitizing hyperproliferative or otherwise cancerous cells (e.g., resistant cells) to the cancer therapy.

The term "synergistic effect" refers to the combined effect of two or more anti- immune checkpoint agents can be greater than the sum of the separate effects of the anticancer agents alone.

"Short interfering RNA" (siRNA), also referred to herein as "small interfering RNA" is defined as an agent which functions to inhibit expression of a target biomarker nucleic acid, e.g., by RNAi. An siRNA may be chemically synthesized, may be produced by in vitro transcription, or may be produced within a host cell. In one embodiment, siRNA is a double stranded RNA (dsRNA) molecule of about 15 to about 40 nucleotides in length, preferably about 15 to about 28 nucleotides, more preferably about 19 to about 25 nucleotides in length, and more preferably about 19, 20, 21, or 22 nucleotides in length, and may contain a 3' and/or 5' overhang on each strand having a length of about 0, 1, 2, 3, 4, or 5 nucleotides. The length of the overhang is independent between the two strands, i.e., the length of the overhang on one strand is not dependent on the length of the overhang on the second strand. Preferably the siRNA is capable of promoting RNA interference through degradation or specific post-transcriptional gene silencing (PTGS) of the target messenger RNA (mRNA).

In another embodiment, an siRNA is a small hairpin (also called stem loop) RNA (shRNA). In one embodiment, these shRNAs are composed of a short (e.g., 19-25 nucleotide) antisense strand, followed by a 5-9 nucleotide loop, and the analogous sense strand. Alternatively, the sense strand may precede the nucleotide loop structure and the antisense strand may follow. These shRNAs may be contained in plasmids, retroviruses, and lentiviruses and expressed from, for example, the pol III U6 promoter, or another promoter (see, e.g., Stewart, et al. (2003) RNA 9:493-501 incorporated by reference herein).

RNA interfering agents, e.g., siRNA molecules, may be administered to a patient having or at risk for having cancer, to inhibit expression of a biomarker gene which is overexpressed in cancer and thereby treat, prevent, or inhibit cancer in the subject.

The term "subject" refers to any healthy animal, mammal or human, or any animal, mammal or human afflicted with a cancer, e.g., lung, ovarian, pancreatic, liver, breast, prostate, and colon carcinomas, as well as melanoma and multiple myeloma. The term "subject" is interchangeable with "patient."

The term "survival" includes all of the following: survival until mortality, also known as overall survival (wherein said mortality may be either irrespective of cause or tumor related); "recurrence-free survival" (wherein the term recurrence shall include both localized and distant recurrence); metastasis free survival; disease free survival (wherein the term disease shall include cancer and diseases associated therewith). The length of said survival may be calculated by reference to a defined start point (e.g. time of diagnosis or start of treatment) and end point (e.g. death, recurrence or metastasis). In addition, criteria for efficacy of treatment can be expanded to include response to chemotherapy, probability of survival, probability of metastasis within a given time period, and probability of tumor recurrence.

The term "therapeutic effect" refers to a local or systemic effect in animals, particularly mammals, and more particularly humans, caused by a pharmacologically active substance. The term thus means any substance intended for use in the diagnosis, cure, mitigation, treatment or prevention of disease or in the enhancement of desirable physical or mental development and conditions in an animal or human. The phrase "therapeutically- effective amount" means that amount of such a substance that produces some desired local or systemic effect at a reasonable benefit/risk ratio applicable to any treatment. In certain embodiments, a therapeutically effective amount of a compound will depend on its therapeutic index, solubility, and the like. For example, certain compounds discovered by the methods of the present invention may be administered in a sufficient amount to produce a reasonable benefit/risk ratio applicable to such treatment.

The terms "therapeutically-effective amount" and "effective amount" as used herein means that amount of a compound, material, or composition comprising a compound of the present invention which is effective for producing some desired therapeutic effect in at least a sub-population of cells in an animal at a reasonable benefit/risk ratio applicable to any medical treatment. Toxicity and therapeutic efficacy of subject compounds may be determined by standard pharmaceutical procedures in cell cultures or experimental animals, e.g., for determining the LD50 and the ED50. Compositions that exhibit large therapeutic indices are preferred. In some embodiments, the LD50 (lethal dosage) can be measured and can be, for example, at least 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, 200%, 300%, 400%, 500%, 600%, 700%, 800%, 900%, 1000% or more reduced for the agent relative to no administration of the agent. Similarly, the ED50 (i.e., the concentration which achieves a half-maximal inhibition of symptoms) can be measured and can be, for example, at least 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, 200%, 300%, 400%, 500%, 600%, 700%, 800%, 900%, 1000% or more increased for the agent relative to no administration of the agent. Also, Similarly, the ICso (i.e., the concentration which achieves half-maximal cytotoxic or cytostatic effect on cancer cells) can be measured and can be, for example, at least 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, 200%, 300%, 400%, 500%, 600%, 700%, 800%, 900%, 1000% or more increased for the agent relative to no administration of the agent. In some embodiments, cancer cell growth in an assay can be inhibited by at least about 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or even 100%. In another embodiment, at least about a 10% , 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or even 100% decrease in a solid malignancy can be achieved.

A "transcribed polynucleotide" or "nucleotide transcript" is a polynucleotide (e.g. an mRNA, hnRNA, a cDNA, or an analog of such RNA or cDNA) which is complementary to or homologous with all or a portion of a mature mRNA made by transcription of a biomarker nucleic acid and normal post-transcriptional processing (e.g. splicing), if any, of the RNA transcript, and reverse transcription of the RNA transcript. As used herein, the term "unresponsiveness" includes refractivity of immune cells to stimulation, e.g., stimulation via an activating receptor or a cytokine. Unresponsiveness can occur, e.g., because of exposure to immunosuppressants or exposure to high doses of antigen. As used herein, the term "anergy" or "tolerance" includes refractivity to activating receptor-mediated stimulation. Such refractivity is generally antigen-specific and persists after exposure to the tolerizing antigen has ceased. For example, anergy in T cells (as opposed to unresponsiveness) is characterized by lack of cytokine production, e.g., IL-2. T cell anergy occurs when T cells are exposed to antigen and receive a first signal (a T cell receptor or CD-3 mediated signal) in the absence of a second signal (a costimulatory signal). Under these conditions, reexposure of the cells to the same antigen (even if reexposure occurs in the presence of a costimulatory polypeptide) results in failure to produce cytokines and, thus, failure to proliferate. Anergic T cells can, however, proliferate if cultured with cytokines (e.g., IL-2). For example, T cell anergy can also be observed by the lack of IL-2 production by T lymphocytes as measured by ELIS A or by a proliferation assay using an indicator cell line. Alternatively, a reporter gene construct can be used. For example, anergic T cells fail to initiate IL-2 gene transcription induced by a heterologous promoter under the control of the 5' IL-2 gene enhancer or by a multimer of the API sequence that can be found within the enhancer (Kang et al. (1992) Science 257: 1134).

There is a known and definite correspondence between the amino acid sequence of a particular protein and the nucleotide sequences that can code for the protein, as defined by the genetic code (shown below). Likewise, there is a known and definite correspondence between the nucleotide sequence of a particular nucleic acid and the amino acid sequence encoded by that nucleic acid, as defined by the genetic code.

GENETIC CODE

Alanine (Ala, A) GCA, GCC, GCG, GCT

Arginine (Arg, R) AGA, ACG, CGA, CGC, CGG, CGT

Asparagine (Asn, N) AAC, AAT

Aspartic acid (Asp, D) GAC, GAT

Cysteine (Cys, C) TGC, TGT

Glutamic acid (Glu, E) GAA, GAG

Glutamine (Gin, Q) CAA, CAG

Glycine (Gly, G) GGA, GGC, GGG, GGT

Histidine (His, H) CAC, CAT Isoleucine (He, I) ATA, ATC, ATT

Leucine (Leu, L) CTA, CTC, CTG, CTT, TTA, TTG

Lysine (Lys, K) AAA, AAG

Methionine (Met, M) ATG

Phenylalanine (Phe, Έ) TTC, TTT

Proline (Pro, P) CCA, CCC, CCG, CCT

Serine (Ser, S) AGC, AGT, TCA, TCC, TCG, TCT

Threonine (Thr, T) ACA, ACC, ACG, ACT

Tryptophan (Trp, W) TGG

Tyrosine (Tyr, Y) TAC, TAT

Valine (Val, V) GTA, GTC, GTG, GTT

Termination signal (end) TAA, TAG, TGA

An important and well known feature of the genetic code is its redundancy, whereby, for most of the amino acids used to make proteins, more than one coding nucleotide triplet may be employed (illustrated above). Therefore, a number of different nucleotide sequences may code for a given amino acid sequence. Such nucleotide sequences are considered functionally equivalent since they result in the production of the same amino acid sequence in all organisms (although certain organisms may translate some sequences more efficiently than they do others). Moreover, occasionally, a methylated variant of a purine or pyrimidine may be found in a given nucleotide sequence. Such methylations do not affect the coding relationship between the trinucleotide codon and the corresponding amino acid.

In view of the foregoing, the nucleotide sequence of a DNA or RNA encoding a biomarker nucleic acid (or any portion thereof) can be used to derive the polypeptide amino acid sequence, using the genetic code to translate the DNA or RNA into an amino acid sequence. Likewise, for polypeptide amino acid sequence, corresponding nucleotide sequences that can encode the polypeptide can be deduced from the genetic code (which, because of its redundancy, will produce multiple nucleic acid sequences for any given amino acid sequence). Thus, description and/or disclosure herein of a nucleotide sequence which encodes a polypeptide should be considered to also include description and/or disclosure of the amino acid sequence encoded by the nucleotide sequence. Similarly, description and/or disclosure of a polypeptide amino acid sequence herein should be considered to also include description and/or disclosure of all possible nucleotide sequences that can encode the amino acid sequence.

Finally, nucleic acid and amino acid sequence information for the loci and biomarkers of the present invention (e.g., biomarkers listed in Table 1) are well known in the art and readily available on publicly available databases, such as the National Center for Biotechnology Information (NCBI). For example, exemplary nucleic acid and amino acid sequences derived from publicly available sequence databases are provided below.

Table 1

SEP ID NO: 1 Human PBRM1 Transcript Variant 1 cDNA Sequence

(NM 018313.4)

1 gcggccgcgg ccggaggagc aatagcagca gccgtggcgg ccacggggcg gggcgcggcg 61 gtcggtgacc gcggccgggg ctgcaggcgg cggagcggct ggaagttgga ttccatgggt 121 tccaagagaa gaagagctac ctccccttcc agcagtgtca gcggggactt tgatgatggg 181 caccattctg tgtcaacacc aggcccaagc aggaaaagga ggagactttc caatcttcca 241 actgtagatc ctattgccgt gtgccatgaa ctctataata ccatccgaga ctataaggat 301 gaacagggca gacttctctg tgagctcttc attagggcac caaagcgaag aaatcaacca 361 gactattatg aagtggtttc tcagcccatt gacttgatga aaatccaaca gaaactaaaa 421 atggaagagt atgatgatgt taatttgctg actgctgact tccagcttct ttttaacaat 481 gcaaagtcct attataagcc agattctcct gaatataaag ccgcttgcaa actctgggat 541 ttgtaccttc gaacaagaaa tgagtttgtt cagaaaggag aagcagatga cgaagatgat 601 gatgaagatg ggcaagacaa tcagggcaca gtgactgaag gatcttctcc agcttacttg 661 aaggagatcc tggagcagct tcttgaagcc atagttgtag ctacaaatcc atcaggacgt 721 ctcattagcg aactttttca gaaactgcct tctaaagtgc aatatccaga ttattatgca 781 ataattaagg agcctataga tctcaagacc attgcccaga ggatacagaa tggaagctac 841 aaaagtattc atgcaatggc caaagatata gatctcctcg caaaaaatgc caaaacttat 901 aatgagcctg gctctcaagt attcaaggat gcaaattcaa ttaaaaaaat attttatatg 961 aaaaaggctg aaattgaaca tcatgaaatg gctaagtcaa gtcttcgaat gaggactcca 1021 tccaacttgg ctgcagccag actgacaggt ccttcacaca gtaaaggcag ccttggtgaa 1081 gagagaaatc ccactagcaa gtattaccgt aataaaagag cagtacaagg aggtcgttta 1141 tcagcaatta caatggcact tcaatatggc tcagaaagtg aagaagatgc tgctttagct 1201 gctgcacgct atgaagaggg agagtcagaa gcagaaagca tcacttcctt tatggatgtt 1261 tcaaatcctt tttatcagct ttatgacaca gttaggagtt gtcggaataa ccaagggcag 1321 ctaatagctg aaccttttta ccatttgcct tcaaagaaaa aataccctga ttattaccag 1381 caaattaaaa tgcccatatc actacaacag atccgaacaa aactgaagaa tcaagaatat 1441 gaaactttag atcatttgga gtgtgatctg aatttaatgt ttgaaaatgc caaacgctat 1501 aatgtgccca attcagccat ctacaagcga gttctaaaat tgcagcaagt tatgcaggca 1561 aagaagaaag agcttgccag gagagacgat atcgaggacg gagacagcat gatctcttca 1621 gccacctctg atactggtag tgccaaaaga aaaagtaaaa agaacataag aaagcagcga 1681 atgaaaatct tattcaatgt tgttcttgaa gctcgagagc caggttcagg cagaagactt 1741 tgtgacctat ttatggttaa accatccaaa aaggactatc ctgattatta taaaatcatc 1801 ttggagccaa tggacttgaa aataattgag cataacatcc gcaatgacaa atatgctggt 1861 gaagagggaa tgatagaaga catgaagctg atgttccgga atgccaggca ctataatgag 1921 gagggctccc aggtttataa tgatgcacat atcctggaga agttactcaa ggagaaaagg 1981 aaagagctgg gcccactgcc tgatgatgat gacatggctt ctcccaaact caagctgagt 2041 aggaagagtg gcatttctcc taaaaaatca aaatacatga ctccaatgca gcagaaacta 2101 aatgaggtct atgaagctgt aaagaactat actgataaga ggggtcgccg cctcagtgcc 2161 atatttctga ggcttccctc tagatctgag ttgcctgact actatctgac tattaaaaag 2221 cccatggaca tggaaaaaat tcgaagtcac atgatggcca acaagtacca agatattgac 2281 tctatggttg aggactttgt catgatgttt aataatgcct gtacatacaa tgagccggag 2341 tctttgatct acaaagatgc tcttgttcta cacaaagtcc tgcttgaaac acgcagagac 2401 ctggagggag atgaggactc tcatgtccca aatgtgactt tgctgattca agagcttatc 2461 cacaatcttt ttgtgtcagt catgagtcat caggatgatg agggaagatg ctacagcgat 2521 tctttagcag aaattcctgc tgtggatccc aactttccta acaaaccacc ccttacattt 2581 gacataatta ggaagaatgt tgaaaataat cgctaccgtc ggcttgattt atttcaagag 2641 catatgtttg aagtattgga acgagcaaga aggatgaatc ggacagattc agaaatatat 2701 gaagatgcag tagaacttca gcagtttttt attaaaattc gtgatgaact ctgcaaaaat 2761 ggagagattc ttctttcacc ggcactcagc tataccacaa aacatttgca taatgatgtg 2821 gagaaagaga gaaaggaaaa attgccaaaa gaaatagagg aagataaact aaaacgagaa 2881 gaagaaaaaa gagaagctga aaagagtgaa gattcctctg gtgctgcagg cctctcaggc 2941 ttacatcgca catacagcca ggactgtagc tttaaaaaca gcatgtacca tgttggagat 3001 tacgtctatg tggaacctgc agaggccaac ctacaaccac atatcgtctg tattgaaaga 3061 ctgtgggagg attcagctga aaaagaagtt tttaagagtg actattacaa caaagttcca 3121 gttagtaaaa ttctaggcaa gtgtgtggtc atgtttgtca aggaatactt taagttatgc 3181 ccagaaaact tccgagatga ggatgttttt gtctgtgaat cacggtattc tgccaaaacc 3241 aaatctttta agaaaattaa actgtggacc atgcccatca gctcagtcag gtttgtccct 3301 cgggatgtgc ctctgcctgt ggttcgcgtg gcctctgtat ttgcaaatgc agataaaggt 3361 gatgatgaga agaatacaga caactcagag gacagtcgag ctgaagacaa ttttaacttg 3421 gaaaaggaaa aagaagatgt ccctgtggaa atgtccaatg gtgaaccagg ttgccactac 3481 tttgagcagc tccattacaa tgacatgtgg ctgaaggttg gcgactgtgt cttcatcaag 3541 tcccatggcc tggtgcgtcc tcgtgtgggc agaattgaaa aagtatgggt tcgagatgga 3601 gctgcatatt tttatggccc catcttcatt cacccagaag aaacagagca tgagcccaca 3661 aaaatgttct acaaaaaaga agtatttctg agtaatctgg aagaaacctg ccccatgaca 3721 tgtattctcg gaaagtgtgc tgtgttgtca ttcaaggact tcctctcctg caggccaact 3781 gaaataccag aaaatgacat tctgctttgt gagagccgct acaatgagag cgacaagcag 3841 atgaagaaat tcaaaggatt gaagaggttt tcactctctg ctaaagtggt agatgatgaa 3901 atttactact tcagaaaacc aattgttcct cagaaggagc catcaccttt gctggaaaag 3961 aagatccagt tgctagaagc taaatttgcc gagttagaag gtggagatga tgatattgaa 4021 gagatgggag aagaagatag tgagtctacc ccaaagtctg ccaaaggcag tgcaaagaag 4081 gaaggctcca aacggaaaat caacatgagt ggctacatcc tgttcagcag tgagatgagg 4141 gctgtgatta aggcccaaca cccagactac tctttcgggg agctcagccg cctggtgggg 4201 acagaatgga gaaatcttga gacagccaag aaagcagaat atgaaggcat gatgggtggc 4261 tatccgccag gccttccacc tttgcagggc ccagttgatg gccttgttag catgggcagc 4321 atgcagccac ttcaccctgg ggggcctcca ccccaccatc ttccgccagg tgtgcctggc 4381 ctcccgggca tcccaccacc gggtgtgatg aaccaaggag tggcccctat ggtagggact 4441 ccagcaccag gtggaagtcc atatggacaa caggtgggag ttttggggcc tccagggcag 4501 caggcaccac ctccatatcc cggcccacat ccagctggac cccctgtcat acagcagcca 4561 acaacaccca tgtttgtagc tcccccacca aagacccagc ggcttcttca ctcagaggcc 4621 tacctgaaat acattgaagg actcagtgcg gagtccaaca gcattagcaa gtgggatcag 4681 acactggcag ctcgaagacg cgacgtccat ttgtcgaaag aacaggagag ccgcctaccc 4741 tctcactggc tgaaaagcaa aggggcccac accaccatgg cagatgccct ctggcgcctt 4801 cgagatttga tgctccggga caccctcaac attcgccaag catacaacct agaaaatgtt 4861 taatcacatc attacgtttc ttttatatag aagcataaag agttgtggat cagtagccat 4921 tttagttact gggggtgggg ggaaggaaca aaggaggata atttttattg cattttactg 4981 tacatcacaa ggccattttt atatacggac acttttaata agctatttca atttgtttgt 5041 tatattaagt tgactttatc aaatacacaa agattttttt gcatatgttt ccttcgttta 5101 aaaccagttt cataattggt tgtatatgta gacttggagt tttatctttt tacttgttgc 5161 catggaactg aaaccattag aggtttttgt cttggcttgg ggtttttgtt ttcttggttt 5221 tgggtttttt tatatatata tataaaagaa caaaatgaaa aaaaacacac acacacaaga 5281 gtttacagat tagtttaaat tgataatgaa atgtgaagtt tgtcctagtt tacatcttag 5341 agaggggagt atacttgtgt ttgtttcatg tgcctgaata tcttaagcca ctttctgcaa 5401 aagctgtttc ttacagatga agtgctttct ttgaaaggtg gttatttagg ttttagatgt 5461 ttaatagaca cagcacattt gctctattaa ctcagaggct cactacagaa atatgtaatc 5521 agtgctgtgc atctgtctgc agctaatgta cctcctggac accaggaggg gaaaaagcac 5581 tttttcaatt gtgctgagtt agacatctgt gagttagact atggtgtcag tgatttttgc 5641 agaacacgtg cacaaccctg aggtatgttt aatctaggca ggtacgttta aggatatttt 5701 gatctattta taatgaattc acaatttatg cctataaatt tcagatgatt taaaatttta 5761 aacctgttac attgaaaaac attgaagttc gtcttgaaga aagcattaag gtatgcatgg 5821 aggtgattta tttttaaaca taacacctaa cctaacatgg gtaagagagt atggaactag 5881 atatgagctg tataagaagc ataattgtga acaagtagat tgattgcctt catatacaag 5941 tatgttttag tattccttat ttccttatta tcagatgtat tttttctttt aagtttcaat 6001 gttgttataa ttctcaacca gaaatttaat actttctaaa atatttttta aatttagctt 6061 gtgcttttga attacaggag aagggaatca taatttaata aaacgcttac tagaaagacc 6121 attacagatc ccaaacactt gggtttggtg accctgtctt tcttatatga ccctacaata 6181 aacatttgaa ggcagcatag gatggcagac agtaggaaca ttgtttcact tggcggcatg 6241 tttttgaaac ctgctttata gtaactgggt gattgccatt gtggtagagc ttccactgct 6301 gtttataatc tgagagagtt aatctcagag gatgcttttt tccttttaat ctgctatgaa 6361 tcagtaccca gatgtttaat tactgtactt attaaatcat gagggcaaaa gagtgtagaa 6421 tggaaaaaag tctcttgtat ctagatactt taaatatggg aggcccttta acttaattgc 6481 ctttagtcaa ccactggatt tgaatttgca tcaagtattt taaataatat tgaatttaaa 6541 aaaatgtatt gcagtagtgt gtcagtacct tattgttaaa gtgagtcaga taaatcttca 6601 attcctggct atttgggcaa ttgaatcatc atggactgta taatgcaatc agattatttt 6661 gtttctagac atccttgaat tacaccaaag aacatgaaat ttagttgtgg ttaaattatt 6721 tatttatttc atgcattcat tttatttccc ttaaggtctg gatgagactt ctttggggag 6781 cctctaaaaa aatttttcac tgggggccac gtgggtcatt agaagccaga gctctcctcc 6841 aggctccttc ccagtgccta gaggtgctat aggaaacata gatccagcca ggggcttccc 6901 taaagcagtg cagcaccggc ccagggcatc actagacagg ccctaattaa gtttttttta 6961 aaaagcctgt gtatttattt tagaatcatg tttttctgta tattaacttg ggggatatcg 7021 ttaatattta ggatataaga tttgaggtca gccatcttca aaaaagaaaa aaaaattgac 7081 tcaagaaagt acaagtaaac tatacacctt tttttcataa gttttaggaa ctgtagtaat 7141 gtggcttaga aagtataatg gcctaaatgt tttcaaaatg taagttcctg tggagaagaa 7201 ttgtttatat tgcaaacggg gggactgagg ggaacctgta ggtttaaaac agtatgtttg 7261 tcagccaact gatttaaaag gcctttaact gttttggttg ttgttttttt tttaagccac 7321 tctccccttc ctatgaggaa gaattgagag gggcacctat ttctgtaaaa tccccaaatt 7381 ggtgttgatg attttgagct tgaatgtttt catacctgat taaaacttgg tttattctaa 7441 tttctgtatc atatcatctg aggtttacgt ggtaactagt cttataacat gtatgtatct 7501 tttttttgtt gttcatctaa agctttttaa tccaaataaa tacagagttt gcaaagtgat 7561 ttggattaac caggaaaaaa aaaaaaaaaa aa

SEP ID NO: 2 Human PBRM1 Variant 1 Amino Acid Sequence (NP 060783.3)

1 mgskrrrats psssvsgdfd dghhsvstpg psrkrrrlsn lptvdpiavc helyntirdy 61 kdeqgrllce Ifirapkrrn qpdyyevvsq pidlmkiqqk lkmeeyddvn lltadfqllf 121 nnaksyykpd speykaackl wdlylrtrne fvqkgeadde dddedgqdnq gtvtegsspa 181 ylkeileqll eaivvatnps grliselfqk lpskvqypdy yaiikepidl ktiaqriqng 241 syksihamak didllaknak tynepgsqvf kdansikkif ymkkaeiehh emaksslrmr 301 tpsnlaaarl tgpshskgsl geernptsky yrnkravqgg rlsaitmalq ygseseedaa 361 laaaryeege seaesits fm dvsnpfyqly dtvrscrnnq gqliaepfyh lpskkkypdy 421 yqqikmpisl qqirtklknq eyetldhlec dlnlmfenak rynvpnsaiy krvlklqqvm 481 qakkkelarr ddiedgdsmi ssatsdtgsa krkskknirk qrmkilfnvv learepgsgr 541 rlcdlfmvkp skkdypdyyk iilepmdlki iehnirndky ageegmiedm klmfrnarhy 601 neegsqvynd ahilekllke krkelgplpd dddmaspklk lsrksgispk kskymtpmqq 661 klnevyeavk nytdkrgrrl saiflrlpsr selpdyylti kkpmdmekir shmmankyqd 721 idsmvedfvm mfnnactyne pesliykdal vlhkvlletr rdlegdedsh vpnvtlliqe 781 lihnlfvsvm shqddegrcy sdslaeipav dpnfpnkppl tfdiirknve nnryrrldlf 841 qehmfevler arrmnrtdse iyedavelqq ffikirdelc kngeillspa lsyttkhlhn 901 dvekerkekl pkeieedklk reeekreaek sedssgaagl sglhrtysqd cs fknsmyhv 961 gdyvyvepae anlqphivci erlwedsaek evfksdyynk vpvskilgkc vvmfvkeyfk 1021 Icpenfrded vfvcesrysa ktksfkkikl wtmpissvrf vprdvplpvv rvasvfanad 1081 kgddekntdn sedsraednf nlekekedvp vemsngepgc hyfeqlhynd mwlkvgdcvf 1141 ikshglvrpr vgriekvwvr dgaayfygpi fihpeetehe ptkmfykkev flsnleetcp 1201 mtcilgkcav lsfkdflscr pteipendil lcesrynesd kqmkkfkglk rfslsakvvd 1261 deiyyfrkpi vpqkepspll ekkiqlleak faeleggddd ieemgeedse stpksakgsa 1321 kkegskrkin msgyilfsse mravikaqhp dysfgelsrl vgtewrnlet akkaeyegmm 1381 ggyppglppl qgpvdglvsm gsmqplhpgg ppphhlppgv pglpgipppg vmnqgvapmv 1441 gtpapggspy gqqvgvlgpp gqqapppypg phpagppviq qpttpmfvap ppktqrllhs 1501 eaylkyiegl saesnsiskw dqtlaarrrd vhlskeqesr lpshwlkskg ahttmadalw 1561 rlrdlmlrdt lnirqaynle nv

SEP ID NO: 3 Human PBRM1 Transcript Variant 2 cDNA Sequence

(NM 181042.4)

1 gcggccgggg ctgcaggcgg cggagcggct ggcttgccaa cacttggtgt cacatgtgag 61 cctcccacat gtattcactc tccattccag ctctgtgatt gaactctgct cttattgact 121 agggggcagt tgggcaggca tgcctcattc ctggaattga cagtcattcc taataagttg 181 gattccatgg gttccaagag aagaagagct acctcccctt ccagcagtgt cagcggggac 241 tttgatgatg ggcaccattc tgtgtcaaca ccaggcccaa gcaggaaaag gaggagactt 301 tccaatcttc caactgtaga tcctattgcc gtgtgccatg aactctataa taccatccga 361 gactataagg atgaacaggg cagacttctc tgtgagctct tcattagggc accaaagcga 421 agaaatcaac cagactatta tgaagtggtt tctcagccca ttgacttgat gaaaatccaa 481 cagaaactaa aaatggaaga gtatgatgat gttaatttgc tgactgctga cttccagctt 541 ctttttaaca atgcaaagtc ctattataag ccagattctc ctgaatataa agccgcttgc 601 aaactctggg atttgtacct tcgaacaaga aatgagtttg ttcagaaagg agaagcagat 661 gacgaagatg atgatgaaga tgggcaagac aatcagggca cagtgactga aggatcttct 721 ccagcttact tgaaggagat cctggagcag cttcttgaag ccatagttgt agctacaaat 781 ccatcaggac gtctcattag cgaacttttt cagaaactgc cttctaaagt gcaatatcca 841 gattattatg caataattaa ggagcctata gatctcaaga ccattgccca gaggatacag 901 aatggaagct acaaaagtat tcatgcaatg gccaaagata tagatctcct cgcaaaaaat 961 gccaaaactt ataatgagcc tggctctcaa gtattcaagg atgcaaattc aattaaaaaa 1021 atattttata tgaaaaaggc tgaaattgaa catcatgaaa tggctaagtc aagtcttcga 1081 atgaggactc catccaactt ggctgcagcc agactgacag gtccttcaca cagtaaaggc 1141 agccttggtg aagagagaaa tcccactagc aagtattacc gtaataaaag agcagtacaa 1201 ggaggtcgtt tatcagcaat tacaatggca cttcaatatg gctcagaaag tgaagaagat 1261 gctgctttag ctgctgcacg ctatgaagag ggagagtcag aagcagaaag catcacttcc 1321 tttatggatg tttcaaatcc tttttatcag ctttatgaca cagttaggag ttgtcggaat 1381 aaccaagggc agctaatagc tgaacctttt taccatttgc cttcaaagaa aaaataccct 1441 gattattacc agcaaattaa aatgcccata tcactacaac agatccgaac aaaactgaag 1501 aatcaagaat atgaaacttt agatcatttg gagtgtgatc tgaatttaat gtttgaaaat 1561 gccaaacgct ataatgtgcc caattcagcc atctacaagc gagttctaaa attgcagcaa 1621 gttatgcagg caaagaagaa agagcttgcc aggagagacg atatcgagga cggagacagc 1681 atgatctctt cagccacctc tgatactggt agtgccaaaa gaaaaagtaa aaagaacata 1741 agaaagcagc gaatgaaaat cttattcaat gttgttcttg aagctcgaga gccaggttca 1801 ggcagaagac tttgtgacct atttatggtt aaaccatcca aaaaggacta tcctgattat 1861 tataaaatca tcttggagcc aatggacttg aaaataattg agcataacat ccgcaatgac 1921 aaatatgctg gtgaagaggg aatgatagaa gacatgaagc tgatgttccg gaatgccagg 1981 cactataatg aggagggctc ccaggtttat aatgatgcac atatcctgga gaagttactc 2041 aaggagaaaa ggaaagagct gggcccactg cctgatgatg atgacatggc ttctcccaaa 2101 ctcaagctga gtaggaagag tggcatttct cctaaaaaat caaaatacat gactccaatg 2161 cagcagaaac taaatgaggt ctatgaagct gtaaagaact atactgataa gaggggtcgc 2221 cgcctcagtg ccatatttct gaggcttccc tctagatctg agttgcctga ctactatctg 2281 actattaaaa agcccatgga catggaaaaa attcgaagtc acatgatggc caacaagtac 2341 caagatattg actctatggt tgaggacttt gtcatgatgt ttaataatgc ctgtacatac 2401 aatgagccgg agtctttgat ctacaaagat gctcttgttc tacacaaagt cctgcttgaa 2461 acacgcagag acctggaggg agatgaggac tctcatgtcc caaatgtgac tttgctgatt 2521 caagagctta tccacaatct ttttgtgtca gtcatgagtc atcaggatga tgagggaaga 2581 tgctacagcg attctttagc agaaattcct gctgtggatc ccaactttcc taacaaacca 2641 ccccttacat ttgacataat taggaagaat gttgaaaata atcgctaccg tcggcttgat 2701 ttatttcaag agcatatgtt tgaagtattg gaacgagcaa gaaggatgaa tcggacagat 2761 tcagaaatat atgaagatgc agtagaactt cagcagtttt ttattaaaat tcgtgatgaa 2821 ctctgcaaaa atggagagat tcttctttca ccggcactca gctataccac aaaacatttg 2881 cataatgatg tggagaaaga gagaaaggaa aaattgccaa aagaaataga ggaagataaa 2941 ctaaaacgag aagaagaaaa aagagaagct gaaaagagtg aagattcctc tggtgctgca 3001 ggcctctcag gcttacatcg cacatacagc caggactgta gctttaaaaa cagcatgtac 3061 catgttggag attacgtcta tgtggaacct gcagaggcca acctacaacc acatatcgtc 3121 tgtattgaaa gactgtggga ggattcagct ggtgaaaaat ggttgtatgg ctgttggttt 3181 taccgaccaa atgaaacatt ccacctggct acacgaaaat ttctagaaaa agaagttttt 3241 aagagtgact attacaacaa agttccagtt agtaaaattc taggcaagtg tgtggtcatg 3301 tttgtcaagg aatactttaa gttatgccca gaaaacttcc gagatgagga tgtttttgtc 3361 tgtgaatcac ggtattctgc caaaaccaaa tcttttaaga aaattaaact gtggaccatg 3421 cccatcagct cagtcaggtt tgtccctcgg gatgtgcctc tgcctgtggt tcgcgtggcc 3481 tctgtatttg caaatgcaga taaaggtgat gatgagaaga atacagacaa ctcagaggac 3541 agtcgagctg aagacaattt taacttggaa aaggaaaaag aagatgtccc tgtggaaatg 3601 tccaatggtg aaccaggttg ccactacttt gagcagctcc attacaatga catgtggctg 3661 aaggttggcg actgtgtctt catcaagtcc catggcctgg tgcgtcctcg tgtgggcaga 3721 attgaaaaag tatgggttcg agatggagct gcatattttt atggccccat cttcattcac

3781 ccagaagaaa cagagcatga gcccacaaaa atgttctaca aaaaagaagt atttctgagt

3841 aatctggaag aaacctgccc catgacatgt attctcggaa agtgtgctgt gttgtcattc

3901 aaggacttcc tctcctgcag gccaactgaa ataccagaaa atgacattct gctttgtgag

3961 agccgctaca atgagagcga caagcagatg aagaaattca aaggattgaa gaggttttca

4021 ctctctgcta aagtggtaga tgatgaaatt tactacttca gaaaaccaat tgttcctcag

4081 aaggagccat cacctttgct ggaaaagaag atccagttgc tagaagctaa atttgccgag

4141 ttagaaggtg gagatgatga tattgaagag atgggagaag aagatagtga ggtcattgaa

4201 cctccttctc tacctcagct tcagaccccc ctggccagtg agctggacct catgccctac

4261 acacccccac agtctacccc aaagtctgcc aaaggcagtg caaagaagga aggctccaaa

4321 cggaaaatca acatgagtgg ctacatcctg ttcagcagtg agatgagggc tgtgattaag

4381 gcccaacacc cagactactc tttcggggag ctcagccgcc tggtggggac agaatggaga

4441 aatcttgaga cagccaagaa agcagaatat gaaggtgtga tgaaccaagg agtggcccct

4501 atggtaggga ctccagcacc aggtggaagt ccatatggac aacaggtggg agttttgggg

4561 cctccagggc agcaggcacc acctccatat cccggcccac atccagctgg accccctgtc

4621 atacagcagc caacaacacc catgtttgta gctcccccac caaagaccca gcggcttctt

4681 cactcagagg cctacctgaa atacattgaa ggactcagtg cggagtccaa cagcattagc

4741 aagtgggatc agacactggc agctcgaaga cgcgacgtcc atttgtcgaa agaacaggag

4801 agccgcctac cctctcactg gctgaaaagc aaaggggccc acaccaccat ggcagatgcc

4861 ctctggcgcc ttcgagattt gatgctccgg gacaccctca acattcgcca agcatacaac

4921 ctagaaaatg tttaatcaca tcattacgtt tcttttatat agaagcataa agagttgtgg

4981 atcagtagcc attttagtta ctgggggtgg ggggaaggaa caaaggagga taatttttat

5041 tgcattttac tgtacatcac aaggccattt ttatatacgg acacttttaa taagctattt

5101 caatttgttt gttatattaa gttgacttta tcaaatacac aaagattttt ttgcatatgt

5161 ttccttcgtt taaaaccagt ttcataattg gttgtatatg tagacttgga gttttatctt

5221 tttacttgtt gccatggaac tgaaaccatt agaggttttt gtcttggctt ggggtttttg

5281 ttttcttggt tttgggtttt tttatatata tatataaaag aacaaaatga aaaaaaacac

5341 acacacacaa gagtttacag attagtttaa attgataatg aaatgtgaag tttgtcctag

5401 tttacatctt agagagggga gtatacttgt gtttgtttca tgtgcctgaa tatcttaagc

5461 cactttctgc aaaagctgtt tcttacagat gaagtgcttt ctttgaaagg tggttattta

5521 ggttttagat gtttaataga cacagcacat ttgctctatt aactcagagg ctcactacag

5581 aaatatgtaa tcagtgctgt gcatctgtct gcagctaatg tacctcctgg acaccaggag

5641 gggaaaaagc actttttcaa ttgtgctgag ttagacatct gtgagttaga ctatggtgtc

5701 agtgattttt gcagaacacg tgcacaaccc tgaggtatgt ttaatctagg caggtacgtt

5761 taaggatatt ttgatctatt tataatgaat tcacaattta tgcctataaa tttcagatga

5821 tttaaaattt taaacctgtt acattgaaaa acattgaagt tcgtcttgaa gaaagcatta

5881 aggtatgcat ggaggtgatt tatttttaaa cataacacct aacctaacat gggtaagaga

5941 gtatggaact agatatgagc tgtataagaa gcataattgt gaacaagtag attgattgcc

6001 ttcatataca agtatgtttt agtattcctt atttccttat tatcagatgt attttttctt

6061 ttaagtttca atgttgttat aattctcaac cagaaattta atactttcta aaatattttt

6121 taaatttagc ttgtgctttt gaattacagg agaagggaat cataatttaa taaaacgctt

6181 actagaaaga ccattacaga tcccaaacac ttgggtttgg tgaccctgtc tttcttatat

6241 gaccctacaa taaacatttg aaggcagcat aggatggcag acagtaggaa cattgtttca

6301 cttggcggca tgtttttgaa acctgcttta tagtaactgg gtgattgcca ttgtggtaga

6361 gcttccactg ctgtttataa tctgagagag ttaatctcag aggatgcttt tttcctttta

6421 atctgctatg aatcagtacc cagatgttta attactgtac ttattaaatc atgagggcaa

6481 aagagtgtag aatggaaaaa agtctcttgt atctagatac tttaaatatg ggaggccctt

6541 taacttaatt gcctttagtc aaccactgga tttgaatttg catcaagtat tttaaataat

6601 attgaattta aaaaaatgta ttgcagtagt gtgtcagtac cttattgtta aagtgagtca

6661 gataaatctt caattcctgg ctatttgggc aattgaatca tcatggactg tataatgcaa

6721 tcagattatt ttgtttctag acatccttga attacaccaa agaacatgaa atttagttgt

6781 ggttaaatta tttatttatt tcatgcattc attttatttc ccttaaggtc tggatgagac

6841 ttctttgggg agcctctaaa aaaatttttc actgggggcc acgtgggtca ttagaagcca

6901 gagctctcct ccaggctcct tcccagtgcc tagaggtgct ataggaaaca tagatccagc

6961 caggggcttc cctaaagcag tgcagcaccg gcccagggca tcactagaca ggccctaatt

7021 aagttttttt taaaaagcct gtgtatttat tttagaatca tgtttttctg tatattaact

7081 tgggggatat cgttaatatt taggatataa gatttgaggt cagccatctt caaaaaagaa

7141 aaaaaaattg actcaagaaa gtacaagtaa actatacacc tttttttcat aagttttagg

7201 aactgtagta atgtggctta gaaagtataa tggcctaaat gttttcaaaa tgtaagttcc

7261 tgtggagaag aattgtttat attgcaaacg gggggactga ggggaacctg taggtttaaa

7321 acagtatgtt tgtcagccaa ctgatttaaa aggcctttaa ctgttttggt tgttgttttt 7381 tttttaagcc actctcccct tcctatgagg aagaattgag aggggcacct atttctgtaa

7441 aatccccaaa ttggtgttga tgattttgag cttgaatgtt ttcatacctg attaaaactt

7501 ggtttattct aatttctgta tcatatcatc tgaggtttac gtggtaacta gtcttataac

7561 atgtatgtat cttttttttg ttgttcatct aaagcttttt aatccaaat

SEP ID NO: 4 Human PBRM1 Variant 2 Amino Acid Sequence (NP 851385.1)

1 mgskrrrats psssvsgdfd dghhsvstpg psrkrrrlsn lptvdpiavc helyntirdy 61 kdeqgrllce Ifirapkrrn qpdyyevvsq pidlmkiqqk lkmeeyddvn lltadfqllf 121 nnaksyykpd speykaackl wdlylrtrne fvqkgeadde dddedgqdnq gtvtegsspa 181 ylkeileqll eaivvatnps grliselfqk lpskvqypdy yaiikepidl ktiaqriqng 241 syksihamak didllaknak tynepgsqvf kdansikkif ymkkaeiehh emaksslrmr 301 tpsnlaaarl tgpshskgsl geernptsky yrnkravqgg rlsaitmalq ygseseedaa 361 laaaryeege seaesits fm dvsnpfyqly dtvrscrnnq gqliaepfyh lpskkkypdy 421 yqqikmpisl qqirtklknq eyetldhlec dlnlmfenak rynvpnsaiy krvlklqqvm 481 qakkkelarr ddiedgdsmi ssatsdtgsa krkskknirk qrmkilfnvv learepgsgr 541 rlcdlfmvkp skkdypdyyk iilepmdlki iehnirndky ageegmiedm klmfrnarhy 601 neegsqvynd ahilekllke krkelgplpd dddmaspklk lsrksgispk kskymtpmqq 661 klnevyeavk nytdkrgrrl saiflrlpsr selpdyylti kkpmdmekir shmmankyqd 721 idsmvedfvm mfnnactyne pesliykdal vlhkvlletr rdlegdedsh vpnvtlliqe 781 lihnlfvsvm shqddegrcy sdslaeipav dpnfpnkppl tfdiirknve nnryrrldlf 841 qehmfevler arrmnrtdse iyedavelqq ffikirdelc kngeillspa lsyttkhlhn 901 dvekerkekl pkeieedklk reeekreaek sedssgaagl sglhrtysqd cs fknsmyhv 961 gdyvyvepae anlqphivci erlwedsage kwlygcwfyr pnetfhlatr kflekevfks 1021 dyynkvpvsk ilgkcvvmfv keyfklcpen frdedvfvce srysaktksf kkiklwtmpi 1081 ssvrfvprdv plpvvrvasv fanadkgdde kntdnsedsr aednfnleke kedvpvemsn 1141 gepgchyfeq lhyndmwlkv gdcvfikshg lvrprvgrie kvwvrdgaay fygpifihpe 1201 eteheptkmf ykkevflsnl eetcpmtcil gkcavls fkd flscrpteip endillcesr 1261 ynesdkqmkk fkglkrfsis akvvddeiyy frkpivpqke pspllekkiq lleakfaele 1321 ggdddieemg eedseviepp slpqlqtpla seldlmpytp pqstpksakg sakkegskrk 1381 inmsgyilfs semravikaq hpdys fgels rlvgtewrnl etakkaeyeg vmnqgvapmv 1441 gtpapggspy gqqvgvlgpp gqqapppypg phpagppviq qpttpmfvap ppktqrllhs 1501 eaylkyiegl saesnsiskw dqtlaarrrd vhlskeqesr lpshwlkskg ahttmadalw 1561 rlrdlmlrdt lnirqaynle nv

SEP ID NO: 5 Mouse PBRM1 cDNA Sequence (NM 001081251.1)

1 ggatttacgg cagcactggg aggggtgagg gcggtgaggg cggcgggtgc cggagagacg

61 gccgcggcca gaggagcgct agcagccgtg gcggccacgg ggcggggctc ggcggtcggg

121 gaccgcagcc ggggctgcag gcggcggagc ggcgggcttg ccaacacttg gtgtcacatg

181 tgagcctccc acatgtgtgc actctccatt ccagctctgt gattgaactc tgctcttatt

241 gactaggggg cacttgggca ggcatgcttc attcctggag ttgacagtca tttcataaga

301 agttggattc catgggttcc aagagaagaa gagccacctc tccttccagc agtgtcagtg

361 gagactttga tgacgggcac cattctgtgc ctacaccagg cccaagcagg aaaaggagaa

421 gactgtccaa tcttccaact gtagatccta ttgctgtgtg ccatgaactc tataacacca

481 tccgagacta taaggatgaa cagggcagac tcctctgtga gctgttcatt agggctccaa

541 agcggagaaa tcaaccagac tattatgaag tggtttctca gcccattgac ttgatgaaaa

601 tccaacagaa acttaaaatg gaagagtatg atgatgttaa tctactgact gctgacttcc

661 agctgctttt taacaatgca aaggcctact ataagccaga ttcccctgag tataaagctg

721 cttgtaaact ctgggatttg taccttcgaa caagaaatga gtttgttcag aaaggagaag

781 cagacgatga agatgatgac gaagatgggc aagacaatca aggcacactg gctgacggct

841 cttctccagg ttatctgaag gagatcctgg agcagcttct tgaagccata gttgtagcca

901 caaatccatc aggacggctc atcagtgaac tttttcagaa actgccttcc aaagtgcaat

961 atccagacta ttatgcaata attaaggaac ctatagatct caagaccatt gctcagagga

1021 tacagaatgg aagctacaaa agtatacacg caatggccaa agatatagat cttctagcaa

1081 aaaatgccaa aacatacaat gagcctgggt ctcaagtatt caaggatgcc aattcgatta

1141 aaaaaatatt ttatatgaaa aaggcagaaa ttgaacatca tgaaatgact aaatcaagtc

1201 ttcgaataag gactgcatca aatttggctg cagccaggct gacaggtcct tcgcacaata

1261 aaagcagcct tggtgaagaa agaaacccca ctagcaagta ttaccgtaat aaaagagcag

1321 tccaaggggg tcgcttgtca gcaattacca tggcacttca gtatggatca gagagtgaag

1381 aggacgctgc tttagctgct gcacgctatg aagaagggga atctgaagca gagagcatca 1441 cttccttcat ggacgtttcc aacccctttc atcagcttta cgacacagtt aggagctgta 1501 ggaatcacca agggcagctc atagctgaac ctttcttcca tttgccttca aagaaaaaat 1561 acccagatta ttatcagcaa attaaaatgc ccatatcact tcaacagatc agaacaaagc 1621 taaagaacca agaatatgaa actttagatc atttggagtg tgatctgaat ttaatgtttg 1681 aaaatgccaa acgttataac gttcccaatt cagccatcta taagcgagtt ctaaaactgc 1741 agcaagtcat gcaggcaaag aagaaggagc ttgcgaggag agatgacatt gaggacggag 1801 acagcatgat ctcctcagcc acttctgaca ctggtagtgc caaaaggaaa aggaatactc 1861 atgacagtga gatgttgggt ctcaggaggc tatccagtaa aaagaacata agaaaacagc 1921 gaatgaaaat tttattcaat gttgttcttg aagctcgaga gccaggttca ggcagaagac 1981 tttgcgatct atttatggtt aagccatcca agaaggacta tcctgattat tataaaatca 2041 tcttagagcc aatggacctg aaaataattg agcataacat ccgaaatgac aaatatgcag 2101 gtgaagaagg aatgatggaa gacatgaaac tcatgttccg caatgccagg cactacaatg 2161 aggagggctc ccaggtatac aatgatgccc atatcctgga gaagttactc aaagataaaa 2221 ggaaagagct gggccctctg cctgatgatg atgacatggc ttctcccaaa cttaaattga 2281 gtaggaagag tggtgtttct cctaagaaat caaagtacat gactccaatg cagcagaaac 2341 tgaatgaagt gtatgaagct gtaaagaact atactgataa gaggggtcgc cgccttagtg 2401 ctatatttct aagactcccc tctagatcag agctgcctga ctactacctg accattaaaa 2461 agcccatgga catggaaaaa attcgaagtc acatgatggc aaacaagtac caagacatag 2521 attctatggt agaggacttt gtcatgatgt ttaataatgc ctgtacctac aatgaaccag 2581 agtctttgat ctacaaagat gcccttgtac tgcataaagt cctccttgag actcggagag 2641 acctggaggg agatgaggat tctcatgtcc ctaatgtgac gttgctgatt caagagctca 2701 tccataacct ttttgtgtca gtcatgagtc atcaggatga cgaagggagg tgttacagcg 2761 actccttagc agaaattcct gctgtggatc ccaactctcc caataaacct ccccttacat 2821 ttgacattat caggaaaaat gttgaaagta atcggtatcg gcgacttgat ttatttcagg 2881 agcatatgtt tgaagtattg gaacgggcaa gaaggatgaa ccggacagat tccgaaatat 2941 atgaggatgc tgtagaactt cagcagtttt ttattagaat tcgtgatgaa ctctgcaaaa 3001 atggagagat ccttctttct ccagcactca gctataccac aaaacacttg cataacgatg 3061 tggaaaaaga aaaaaaggaa aaattgccta aagaaataga ggaagataaa ctaaaacgcg 3121 aagaagaaaa aagagaagct gaaaaaagtg aagattcctc aggtactaca ggcctctcag 3181 gcttacatcg tacatacagc caggactgca gctttaagaa cagcatgtat catgtcggag 3241 attatgtcta tgttgaacct gcggaggcca atctacaacc acatatagtg tgtattgaga 3301 gactgtggga ggattcagct ggtgaaaaat ggttgtacgg ctgttggttt tatcggccaa 3361 atgaaacatt ccatttggct acacgaaaat ttctagaaaa agaagttttt aagagtgact 3421 actacaataa agtacctgtt agtaaaattc taggcaaatg tgtagtcatg tttgtcaagg 3481 aatactttaa attatgtcca gaaaactttc gcgatgagga tgtttttgtc tgtgaatcga 3541 ggtattctgc caaaaccaaa tcttttaaga aaattaaact gtggaccatg cccatcagtt 3601 cagttagatt tgtccctcgg gatgtgcctt tgcctgtggt ccgagtggcc tctgtgtttg 3661 caaatgcaga taaaggggat gatgagaaga atacagacaa ctcagatgac aatagagctg 3721 aagacaattt taacttggaa aaggaaaaag aagatgttcc tgtggagatg tccaatggtg 3781 agccaggttg ccactacttt gagcagcttc ggtacaatga catgtggctg aaggttggtg 3841 attgtgtctt catcaaatcc cacggcttgg tgcgccctcg tgtgggcaga attgagaaag 3901 tatgggtccg agatggagct gcatattttt atggccctat cttcattcat ccagaagaaa 3961 cagaacatga gcccacaaaa atgttctaca aaaaagaagt gtttctgagt aatctggaag 4021 agacctgccc tatgagttgt attctgggga aatgtgcagt gctgtcattc aaggacttcc 4081 tctcctgcag gccaactgaa ataccagaaa atgacattct gctttgtgag agccgctata 4141 atgagagtga caagcagatg aagaagttca agggtttgaa gaggttttca ctctctgcta 4201 aagttgtaga tgatgaaatc tactacttca gaaaaccaat cattcctcag aaggaaccct 4261 cacctttgtt agaaaagaag atacaattgc tagaagctaa atttgcagag ttagaaggag 4321 gagatgatga tattgaggag atgggagaag aggatagtga agtcattgaa gctccatctc 4381 tacctcaact gcagacaccc ctggccaatg agttggacct catgccctat acacccccac 4441 agtctacccc aaagtctgcc aaaggcagtg caaagaagga aagttctaaa cgaaaaatca 4501 acatgagtgg ctacattttg ttcagcagtg aaatgagagc tgtgattaaa gcccagcacc 4561 cagactactc ttttggggag ctcagcagac tggtggggac agaatggaga aaccttgaaa 4621 cagccaagaa agcagaatat gaagagcggg cagctaaagt tgctgagcag caggagagag 4681 agcgagcagc acagcaacag cagccgagtg cttctccccg agcaggcacc cctgtggggg 4741 ctctcatggg ggtggtgcca ccaccaacac caatggggat gctcaatcag cagttgacac 4801 ctgttgcagg catgatgggt ggctatccgc caggccttcc acctttgcag ggcccagttg 4861 atggccttgt tagcatgggc agcatgcagc cacttcaccc tggggggcct ccacctcacc 4921 atcttccgcc aggtgtgcct ggcctcccag gcatcccacc accgggtgtg atgaatcaag 4981 gagtagcccc catggtaggg actccagcac caggtggaag tccgtatgga caacaggtag 5041 gagttttggg acctccaggg cagcaggcac cacctccata tcctggtcct catccagctg 5101 gcccccctgt catacagcag ccaacaacgc ccatgtttgt ggctccccca ccaaagaccc 5161 aaaggcttct ccactcagag gcctacctga aatacattga aggactcagt gctgaatcca 5221 acagcattag caagtgggac caaactttgg cagctcgaag acgggatgtc catttgtcca 5281 aagaacagga gagccgccta ccttctcact ggctcaaaag taaaggggca cacaccacca 5341 tggcagatgc cctctggcgc ctacgggatt taatgcttcg agacactctc aacatccgac 5401 aggcatacaa cctagaaaat gtttaatcac atcactgttt cttctgtgga agcaaagagt 5461 tgtggagcgg tagccatttt agttactggg gtgggaggga ggaacaaagg atgataattt 5521 ttattgcatt ttattgtaca tcacacagcc atttttatat aaggacactt ttaataagct 5581 atttcaaatt tggttttgtt acattaagtt gactatcaaa tacacaaaag attttttttg 5641 catatgtttc ctttgtttaa aaccagtttc ataattggtt atatatagta atagttttat 5701 ctttacttgt taaaggactt aaatcatcaa aggttttggc ttggcttagg gttttcgttt 5761 tcttttttat aaatatatat tatatatata tacacatata aaagaaaaaa tgaaaaaaaa 5821 gtttacaaat ttaagttgac aatgaaatgt gaagttggtc ctagtttaca tcttagagga 5881 atgtatatgt atgttttaca tgcctaaata tctgcaggtt ttcttacagg taaagcgaag 5941 tgctttgaaa agtttagatt atacatgtgt gacagatgcg gcatatttgc tctattaaca 6001 cagaggctta ctatagaaat ctaaagtcaa tgctgtacat ccatccagtt agtgtaactg 6061 aagggaaatg taactttgtg ctgagttaga catctgtatt gtcagtgatt cttgtagaat 6121 atgtgctcag atctgagtta tatttagttt tggaaggtaa gttgaagagt acttttgatc 6181 agtttatgat tcagtttatg attttagttt ttgccttcat gttatacatt tatgatttga 6241 aactgtacat ctgttacctt gaaaaacatt gaagaaagta ctgaagtgtg catggaggtg 6301 gtttaagcat aatacttaac ccaagaaaga gtgtaagtgg acacaagctg tgcctgcaca 6361 tagctgtgca gggtagactg cctacataca catggccggg attctttatt tccttgttat 6421 caattatagt gctttgtttg tttcagggtt ggaattctca accagaaata atactttcta 6481 aaatatttta aaattcagct tgtgctttgg attatagaag gaaattatac tttaagaaaa 6541 tgttcacaaa aaaaaaaaaa aaaaaaggac tattacagat cccaatactt ggatttggtg 6601 accttgtctt tctttctttt cttgagacat ggtcctacta ccaaccctgg ctggactgga 6661 gctcagtgta tagaccaggc tagtctcaaa ctctgcctct tcctcccaag tgctgggatt 6721 aagggcaggt accatagtgc tcagcaacca caaccctgtc tttccaacac ggccctagcg 6781 taagcactga ggcagtgtgc agtgctcagg cagcagcaaa catttcccgg gggtggtttt 6841 gaacctgctt gggtggttgt gtggtgctga cgctgccact gccctgttgt tcattgagaa 6901 tgattgttaa atgacactct tcctttagaa tataacggat cagtactcat gtttaattgc 6961 catgcttaat aaatcatgag aacaaaagag tatagaatgg aaagcattcc ctggtagcta 7021 ctttaaatac aggagccctg taacttaata ccagtagtca accactggat ctcagttttc 7081 atcaagtatt ttaaataaat aatcttaaat tttaaaatac gtactgcaga gtatgccagt 7141 atcttattgt taaaactgaa tcaaataaat cttcgattcc tggttatttg gaccattgac 7201 tcatcatgga ctatataatg taataagatt cttttctctt aaggtatcct tgaattacac 7261 caaagaacca gaaacttaat tttggttaaa ttatttattt atttcatgca ttaattttct 7321 ttttcttttt aaaggtttag atgaggctcc ttagggagtc tctaaaaccg cttcactatc 7381 agcaaccagg agtactagaa gccagagcac tcttcctcct ggctcctccc cagtgctcta 7441 gtgctgtagg aaccaagagc cagccccagg ttccccgagg cagtaaaaat ccagcacagg 7501 gggctgtgtc cctaaggcaa gccctgatta cctttaaaaa aaaccaaaaa aacaaacaaa 7561 aaaaaaaaac ctaattaact aaagcattta aggcactatt tattttagaa tcatgctttt 7621 gaagagcatc agtgattact tagggtgtaa tatgtaaaga tcagacatct ccaaaaacag 7681 aaaaagtaca agtaaacaac acactttctc atgactttta agaactgtag taatgtggct 7741 taggaaatat aatggcctaa ttgttttcaa aatgtaagtt cctgtgaaga attttgttta 7801 tattgggttg gggacctata ggtttaaaat agaatgtcag tcagctgact taaaaaacat 7861 tggttttact aagtctgcct tccccttcta aggaagaact gagtgggtaa gggacaggtg 7921 tgtaaaatct ccaaatggat gttacagctt tcagcttgaa cgtttgtttc cagacctgat 7981 taaaatttgg tttattctaa tttctgtact atatcatctg aggttttaag tggtaactgg 8041 ttctatacca tgtatgtatc atatgtttgt tcatcaaagc tttttaatcc aaataaaaac 8101 aacagtttgc aaagtga

SEP ID NO: 6 Mouse PBRM1 Amino Acid Sequence (NP 001074720.1)

1 mgskrrrats psssvsgdfd dghhsvptpg psrkrrrlsn lptvdpiavc helyntirdy

61 kdeqgrllce lfirapkrrn qpdyyevvsq pidlmkiqqk lkmeeyddvn lltadfqllf

121 nnakayykpd speykaackl wdlylrtrne fvqkgeadde dddedgqdnq gtladgsspg

181 ylkeileqll eaivvatnps grliselfqk lpskvqypdy yaiikepidl ktiaqriqng

241 syksihamak didllaknak tynepgsqvf kdansikkif ymkkaeiehh emtksslrir

301 tasnlaaarl tgpshnkssl geernptsky yrnkravqgg rlsaitmalq ygseseedaa 361 laaaryeege seaesitsfm dvsnpfhqly dtvrscrnhq gqliaepffh lpskkkypdy 421 yqqikmpisl qqirtklknq eyetldhlec dlnlmfenak rynvpnsaiy krvlklqqvm 481 qakkkelarr ddiedgdsmi ssatsdtgsa krkrnthdse mlglrrlssk knirkqrmki 541 lfnvvleare pgsgrrlcdl fmvkpskkdy pdyykiilep mdlkiiehni rndkyageeg 601 mmedmklmfr narhyneegs qvyndahile kllkdkrkel gplpddddma spklklsrks 661 gvspkkskym tpmqqklnev yeavknytdk rgrrlsaifl rlpsrselpd yyltikkpmd 721 mekirshmma nkyqdidsmv edfvmmfnna ctynepesli ykdalvlhkv lletrrdleg 781 dedshvpnvt lliqelihnl fvsvmshqdd egrcysdsla eipavdpnsp nkppltfdii 841 rknvesnryr rldlfqehmf evlerarrmn rtdseiyeda velqqffiri rdelckngei 901 llspalsytt khlhndveke kkeklpkeie edklkreeek reaeksedss gttglsglhr 961 tysqdcsfkn smyhvgdyvy vepaeanlqp hivcierlwe dsagekwlyg cwfyrpnetf 1021 hlatrkflek evfksdyynk vpvskilgkc vvmfvkeyfk lcpenfrded vfvcesrysa 1081 ktksfkkikl wtmpissvrf vprdvplpvv rvasvfanad kgddekntdn sddnraednf 1141 nlekekedvp vemsngepgc hyfeqlrynd mwlkvgdcvf ikshglvrpr vgriekvwvr 1201 dgaayfygpi fihpeetehe ptkmfykkev flsnleetcp mscilgkcav lsfkdflscr 1261 pteipendil lcesrynesd kqmkkfkglk rfslsakvvd deiyyfrkpi ipqkepspll 1321 ekkiqlleak faeleggddd ieemgeedse vieapslpql qtplaneldl mpytppqstp 1381 ksakgsakke sskrkinmsg yilfssemra vikaqhpdys fgelsrlvgt ewrnletakk 1441 aeyeeraakv aeqqereraa qqqqpsaspr agtpvgalmg vvppptpmgm Inqqltpvag 1501 mmggyppglp plqgpvdglv smgsmqplhp ggppphhlpp gvpglpgipp pgvmnqgvap 1561 mvgtpapggs pygqqvgvlg ppgqqapppy pgphpagppv iqqpttpmfv apppktqrll 1621 hseaylkyie glsaesnsis kwdqtlaarr rdvhlskeqe srlpshwlks kgahttmada 1681 lwrlrdlmlr dtlnirqayn lenv * Included in Table 1 are RNA nucleic acid molecules (e.g., thymines replaced with uredines), nucleic acid molecules encoding orthologs of the encoded proteins, as well as DNA or RNA nucleic acid sequences comprising a nucleic acid sequence having at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.5%, or more identity across their full length with the nucleic acid sequence of any SEQ ID NO listed in Table 1, or a portion thereof. Such nucleic acid molecules can have a function of the full-length nucleic acid as described further herein.

* Included in Table 1 are orthologs of the proteins, as well as polypeptide molecules comprising an amino acid sequence having at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.5%, or more identity across their full length with an amino acid sequence of any SEQ ID NO listed in Table 1, or a portion thereof. Such polypeptides can have a function of the full-length polypeptide as described further herein.

* Included in Table 1 is PBRM1, including any PBRM1 cDNA or polypeptide.

* Included in Table 1 are PBRM1 nucleic acid and/or amino acid sequences encoding or representing PBRM1 having reduced or eliminated PBRM1 function (e.g., truncating PBRMl mutations causing encoding of incomplete PBRMl protein). Many of these mutations were found in RCC patients which is insensitive to immune checkpoint therapies. Some exemplary mutations are listed below and are also found in Table 6 filed herewith:

1. Truncating PBRMl alterations in patients training cohort passing whole exome quality control (N=34; genomic positions in the table can be determined from PBRMl_ENST00000337303.4_Nonsense_Mutation_p.E449*|PBRMl_ENST 0000 0394830.3_Nonsense_Mutation_p.E449*|PBRMl_ENST00000409114.3_ Nonsense _Mutation_p.E449*|PBRMl_ENST00000409767.1_Nonsense_Mutation_ p.E449*| PBRMl_ENST00000410007.1_Nonsense_Mutation_p.E449*|PBRMl_ENST 0000 0296302.7_Nonsense_Mutation_p.E449*|PBRMl_ENST00000409057.1_ Nonsense

Mutation_p.E449*; NM_018165.4; or uc003der.2)

Shift ocat Del or

PB 12 3 52678748 52678748 Non C C A p.E2 S 0.15 8 45 1 NA

R 6.0 sens 91* N 0943

Ml 7 e_M P

utati

on

PB 15 3 52620610 52620614 Fra ATTT ATTT p.KI D 0.06 19 26 0 strel

R 5.1 me T T 1087 E 7137 4 ka,

Ml 8 Shift fs L 809 indel

Del ocat or

NA 13 N NA NA NA NA N NA NA NA NA NA 8.8 A A

1

NA 10 N NA NA NA NA N NA NA NA NA NA

0.7 A A

3

PB 94. 3 52613194 52613194 Non c c A p.El S 0.53 53 47 1 NA

R 84 sens 105* N

Ml e_M P

utati

on

PB 14 3 52643375 52643375 Non G G A p.Q8 s 0.28 36 89 1 NA

R 6.6 sens 09* N 8

Ml 9 e_M P

utati

on

PB 11 3 52662964 52662964 Fra A A p.N4 D 0.10 10 82 1 strel

R 1.2 me 63fs E 8695 ka,

Ml 2 Shift L 652 indel

Del ocat or

NA 47. N NA NA NA NA NA NA NA N NA NA NA NA NA 52 A A

PB 13 3 52696272 52696272 Fra T T p.Kl D 0.17 12 58 1 strel

R 0.8 me 35fs E 1428 ka,

Ml 6 Shift L 571 indel

Del ocat or

NA 95. N NA NA NA NA N NA NA NA NA NA 31 A A

PB 26 3 52663052 52663052 Spli c c T S 0.23 25 81 1 NA

R 6.4 ce_S N 5849

Ml ite P

PB 16 3 52643489 52643489 Fra A A p.S8 D 0.40 91 13 1 strel

R 4.3 me 18fs E 2654 5 ka,

Ml 9 Shift L 867 indel

Del ocat or

NA 12 N NA NA NA NA N NA NA NA NA NA 4.8 A A

7

NA 11 N NA NA NA NA N NA NA NA NA NA 5.5 A A

1

PB 17 3 52651277 52651277 Spli C C T S 0.12 6 41 1 NA

R 3.7 ce_S N 766

Ml 8 ite P

PB 67. 3 52621487 52621487 Fra T T p.Nl D 0.46 13 15 1 strel

R 19 me 017f E 4285 ka,

Ml Shift s L 714 indel

Del ocat or

NA 22 N NA NA NA NA N NA NA NA NA NA 1.5 A A

6

NA 12 N NA NA NA NA N NA NA NA NA NA 4.3 A A

5

PB 13 3 52623201 52623201 Fra G G p.D9 D 0.25 15 45 1 strel

R 1.6 me 65fs E ka,

Ml 3 Shift L indel

Del ocat or

11 NA 62. N NA NA NA NA N NA NA NA NA NA 1 73 A A

1

11 PB 89. 3 52623120 52623120 Fra G G p.I9 D 0.55 55 45 1 strel 1 R 9 me 92fs E ka,

0 Ml Shift L indel

Del ocat or

1 PB 13 3 52613062 52613068 Spli ACA ACA D 0.17 37 17 0 strel

62 R 1.1 ce_S CTC CTC E 3708 6 ka

Ml 6 ite A A L 92

1 NA 12 N NA NA NA NA N NA NA NA NA NA

32 0.8 A A

5

1 PB 28. 3 52649455 52649456 Fra T p.H6 I 0.36 8 14 1 strel

20 R 98 me 27fs N 3636 ka,

Ml Shift S 364 indel

Ins ocat or

Patient_id = CA209009_XX (XX: the id in the above table)

2. Tmncating PBRMl alterations in validation cohort (N=28)

PD NA NA NA NA NA NA NA NA NA NA NA NA 018

PD NA NA NA NA NA NA NA NA NA NA NA NA

019

PD NA NA NA NA NA NA NA NA NA NA NA NA

020

PD PB 3 5271 5271 Fram C c p.G2 DEL 0.36 18 32 1 strelk 021 RM 3723 3723 e Sh fs a,

1 ift D indelo el cator

PD NA NA NA NA NA NA NA NA NA NA NA NA

022

PD PB 3 5266 5266 Splic T T A SNP 0.214 9 33 1 NA

023 RM 3053 3053 e_Sit 286

1 e

PD NA NA NA NA NA NA NA NA NA NA NA NA

024

PD PB 3 5259 5259 Fram c c p.Gl DEL 0.154 21 11 1 strelk

025 RM 5829 5829 e Sh 429fs 41176 5 a,

1 ift D 5 indelo el cator

PD NA NA NA NA NA NA NA NA NA NA NA NA

026

RCC NA NA NA NA NA NA NA NA NA NA NA NA

.PD

1.D

NA.

1026

RCC PB 3 5259 5259 Fram c c p.Al DEL 0.133 43 27 1 strelk

.PD RM 5804 5804 e Sh 438fs 54037 9 a,

1.D 1 ift D 3 indelo

NA. el cator

1101

RCC NA NA NA NA NA NA NA NA NA NA NA NA

.PD

1.D

NA.

1137

RCC NA NA NA NA NA NA NA NA NA NA NA NA

.PD

1.D

NA.

944

RCC NA NA NA NA NA NA NA NA NA NA NA NA

.PD

1.D

NA.

949

VA1 PB 3 5264 5264 Fram T T p.K6 DEL 0.06 15 25 not indelo 008 RM 3943 3943 e Sh 19fs 3 eval cator

1 ift D uabl

el e

II. Subjects

In one embodiment, the subject for whom predicted likelihood of efficacy of an immune checkpoint therapy is determined, is a mammal (e.g., mouse, rat, primate, non- human mammal, domestic animal, such as a dog, cat, cow, horse, and the like), and is preferably a human.

In another embodiment of the methods of the present invention, the subject has not undergone treatment, such as chemotherapy, radiation therapy, targeted therapy, and/or immune checkpoint therapy. In still another embodiment, the subject has undergone treatment, such as chemotherapy, radiation therapy, targeted therapy, and/or immune checkpoint therapy.

In certain embodiments, the subject has had surgery to remove cancerous or precancerous tissue. In other embodiments, the cancerous tissue has not been removed, e.g., the cancerous tissue may be located in an inoperable region of the body, such as in a tissue that is essential for life, or in a region where a surgical procedure would cause considerable risk of harm to the patient.

The methods of the present invention can be used to determine the responsiveness to anti-immune checkpoint therapies of renal cell carcinomas, particularly because kidney cancers are genomically different from many cancers according to cancer-related mutational load and composition. However, as described in herein, the methods of the present invention can, in certain embodiments, be applied to cancers other than renal cell carcinoma. In one embodiment, the cancers are solid tumors, such as lung cancer, melanoma, and/or renal cell carcinoma. In another embodiment, the cancer is an epithelial cancer such as, but not limited to, brain cancer (e.g., glioblastomas) bladder cancer, breast cancer, cervical cancer, colon cancer, gynecologic cancers, renal cancer, laryngeal cancer, lung cancer, oral cancer, head and neck cancer, ovarian cancer, pancreatic cancer, prostate cancer, or skin cancer. In still other embodiments, the cancer is breast cancer, prostate cancer, lung cancer, or colon cancer. In still other embodiments, the epithelial cancer is non-small-cell lung cancer, nonpapillary renal cell carcinoma, cervical carcinoma, ovarian carcinoma (e.g., serous ovarian carcinoma), or breast carcinoma. The epithelial cancers may be characterized in various other ways including, but not limited to, serous, endometrioid, mucinous, clear cell, brenner, or undifferentiated. In some embodiments, the cancers are mesenchymal tumors, such as sarcoma. III. Sample Collection, Preparation and Separation

In some embodiments, biomarker amount and/or activity measurement(s) in a sample from a subject is compared to a predetermined control (standard) sample. The sample from the subject is typically from a diseased tissue, such as cancer cells or tissues. The control sample can be from the same subject or from a different subject. The control sample is typically a normal, non-diseased sample. However, in some embodiments, such as for staging of disease or for evaluating the efficacy of treatment, the control sample can be from a diseased tissue. The control sample can be a combination of samples from several different subjects. In some embodiments, the biomarker amount and/or activity measurement(s) from a subject is compared to a pre-determined level. This pre-determined level is typically obtained from normal samples. As described herein, a "pre-determined" biomarker amount and/or activity measurement(s) may be a biomarker amount and/or activity measurement(s) used to, by way of example only, evaluate a subject that may be selected for treatment, evaluate a response to an immune checkpoint therapy, and/or evaluate a response to a combination immune checkpoint therapy. A pre-determined biomarker amount and/or activity measurement(s) may be determined in populations of patients with or without cancer. The pre-determined biomarker amount and/or activity measurement(s) can be a single number, equally applicable to every patient, or the pre- determined biomarker amount and/or activity measurement(s) can vary according to specific subpopulations of patients. Age, weight, height, and other factors of a subject may affect the pre-determined biomarker amount and/or activity measurement(s) of the individual. Furthermore, the pre-determined biomarker amount and/or activity can be determined for each subject individually. In one embodiment, the amounts determined and/or compared in a method described herein are based on absolute measurements.

In another embodiment, the amounts determined and/or compared in a method described herein are based on relative measurements, such as ratios (e.g., biomarker copy numbers, level, and/or activity before a treatment vs. after a treatment, such biomarker measurements relative to a spiked or man-made control, such biomarker measurements relative to the expression of a housekeeping gene, and the like). For example, the relative analysis can be based on the ratio of pre-treatment biomarker measurement as compared to post-treatment biomarker measurement. Pre-treatment biomarker measurement can be made at any time prior to initiation of anti-cancer therapy. Post-treatment biomarker measurement can be made at any time after initiation of anti-cancer therapy. In some embodiments, post-treatment biomarker measurements are made 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1, 12, 13, 14, 15, 16, 17, 18, 19, 20 weeks or more after initiation of anti-cancer therapy, and even longer toward indefinitely for continued monitoring. Treatment can comprise anti-cancer therapy, such as a therapeutic regimen comprising an anti-PDl monoclonal antibody (e.g., nivolumab) alone or in combination with other anti-cancer agents, such as anti-PD-Ll/PD-L2 antibodies, anti-VEGF agents (e.g., bevacizumab), agents described in the Examples, Figures, and Tables, or anti-PBRMl agents.

The pre-determined biomarker amount and/or activity measurement(s) can be any suitable standard. For example, the pre-determined biomarker amount and/or activity measurement(s) can be obtained from the same or a different human for whom a patient selection is being assessed. In one embodiment, the pre-determined biomarker amount and/or activity measurement s) can be obtained from a previous assessment of the same patient. In such a manner, the progress of the selection of the patient can be monitored over time. In addition, the control can be obtained from an assessment of another human or multiple humans, e.g., selected groups of humans, if the subject is a human. In such a manner, the extent of the selection of the human for whom selection is being assessed can be compared to suitable other humans, e.g., other humans who are in a similar situation to the human of interest, such as those suffering from similar or the same condition(s) and/or of the same ethnic group.

In some embodiments of the present invention the change of biomarker amount and/or activity measurement s) from the pre-determined level is about 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, or 5.0 fold or greater, or any range in between, inclusive. Such cutoff values apply equally when the measurement is based on relative changes, such as based on the ratio of pre-treatment biomarker measurement as compared to post-treatment biomarker measurement.

Biological samples can be collected from a variety of sources from a patient including a body fluid sample, cell sample, or a tissue sample comprising nucleic acids and/or proteins. "Body fluids" refer to fluids that are excreted or secreted from the body as well as fluids that are normally not (e.g., amniotic fluid, aqueous humor, bile, blood and blood plasma, cerebrospinal fluid, cerumen and earwax, cowper's fluid or pre-ejaculatory fluid, chyle, chyme, stool, female ejaculate, interstitial fluid, intracellular fluid, lymph, menses, breast milk, mucus, pleural fluid, pus, saliva, sebum, semen, serum, sweat, synovial fluid, tears, urine, vaginal lubrication, vitreous humor, vomit). In a preferred embodiment, the subject and/or control sample is selected from the group consisting of cells, cell lines, histological slides, paraffin embedded tissues, biopsies, whole blood, nipple aspirate, serum, plasma, buccal scrape, saliva, cerebrospinal fluid, urine, stool, and bone marrow. In one embodiment, the sample is serum, plasma, or urine. In another

embodiment, the sample is serum.

The samples can be collected from individuals repeatedly over a longitudinal period of time (e.g., once or more on the order of days, weeks, months, annually, biannually, etc.). Obtaining numerous samples from an individual over a period of time can be used to verify results from earlier detections and/or to identify an alteration in biological pattern as a result of, for example, disease progression, drug treatment, etc. For example, subject samples can be taken and monitored every month, every two months, or combinations of one, two, or three month intervals according to the present invention. In addition, the biomarker amount and/or activity measurements of the subject obtained over time can be conveniently compared with each other, as well as with those of normal controls during the monitoring period, thereby providing the subject's own values, as an internal, or personal, control for long-term monitoring.

Sample preparation and separation can involve any of the procedures, depending on the type of sample collected and/or analysis of biomarker measurement(s). Such procedures include, by way of example only, concentration, dilution, adjustment of pH, removal of high abundance polypeptides {e.g., albumin, gamma globulin, and transferrin, etc.), addition of preservatives and calibrants, addition of protease inhibitors, addition of denaturants, desalting of samples, concentration of sample proteins, extraction and purification of lipids.

The sample preparation can also isolate molecules that are bound in non-covalent complexes to other protein {e.g., carrier proteins). This process may isolate those molecules bound to a specific carrier protein {e.g., albumin), or use a more general process, such as the release of bound molecules from all carrier proteins via protein denaturation, for example using an acid, followed by removal of the carrier proteins.

Removal of undesired proteins {e.g. , high abundance, uninformative, or

undetectable proteins) from a sample can be achieved using high affinity reagents, high molecular weight filters, ultracentrifugation and/or electrodialysis. High affinity reagents include antibodies or other reagents {e.g., aptamers) that selectively bind to high abundance proteins. Sample preparation could also include ion exchange chromatography, metal ion affinity chromatography, gel filtration, hydrophobic chromatography, chromatofocusing, adsorption chromatography, isoelectric focusing and related techniques. Molecular weight filters include membranes that separate molecules on the basis of size and molecular weight. Such filters may further employ reverse osmosis, nanofiltration, ultrafiltration and microfiltration.

Ultracentrifugation is a method for removing undesired polypeptides from a sample.

Ultracentrifugation is the centrifugation of a sample at about 15,000-60,000 rpm while monitoring with an optical system the sedimentation (or lack thereof) of particles.

Electrodialysis is a procedure which uses an electromembrane or semipermable membrane in a process in which ions are transported through semi-permeable membranes from one solution to another under the influence of a potential gradient. Since the membranes used in electrodialysis may have the ability to selectively transport ions having positive or negative charge, reject ions of the opposite charge, or to allow species to migrate through a semipermable membrane based on size and charge, it renders electrodialysis useful for concentration, removal, or separation of electrolytes.

Separation and purification in the present invention may include any procedure known in the art, such as capillary electrophoresis (e.g., in capillary or on-chip) or chromatography (e.g., in capillary, column or on a chip). Electrophoresis is a method which can be used to separate ionic molecules under the influence of an electric field.

Electrophoresis can be conducted in a gel, capillary, or in a microchannel on a chip.

Examples of gels used for electrophoresis include starch, acrylamide, polyethylene oxides, agarose, or combinations thereof. A gel can be modified by its cross-linking, addition of detergents, or denaturants, immobilization of enzymes or antibodies (affinity

electrophoresis) or substrates (zymography) and incorporation of a pH gradient. Examples of capillaries used for electrophoresis include capillaries that interface with an electrospray.

Capillary electrophoresis (CE) is preferred for separating complex hydrophilic molecules and highly charged solutes. CE technology can also be implemented on microfluidic chips. Depending on the types of capillary and buffers used, CE can be further segmented into separation techniques such as capillary zone electrophoresis (CZE), capillary isoelectric focusing (CIEF), capillary isotachophoresis (cITP) and capillary electrochromatography (CEC). An embodiment to couple CE techniques to electrospray ionization involves the use of volatile solutions, for example, aqueous mixtures containing a volatile acid and/or base and an organic such as an alcohol or acetonitrile.

Capillary isotachophoresis (cITP) is a technique in which the analytes move through the capillary at a constant speed but are nevertheless separated by their respective mobilities. Capillary zone electrophoresis (CZE), also known as free-solution CE (FSCE), is based on differences in the electrophoretic mobility of the species, determined by the charge on the molecule, and the frictional resistance the molecule encounters during migration which is often directly proportional to the size of the molecule. Capillary isoelectric focusing (CIEF) allows weakly-ionizable amphoteric molecules, to be separated by electrophoresis in a pH gradient. CEC is a hybrid technique between traditional high performance liquid chromatography (UPLC) and CE. Separation and purification techniques used in the present invention include any chromatography procedures known in the art. Chromatography can be based on the differential adsorption and elution of certain analytes or partitioning of analytes between mobile and stationary phases. Different examples of chromatography include, but not limited to, liquid chromatography (LC), gas chromatography (GC), high performance liquid chromatography (HPLC), etc.

IV. Biomarker Nucleic Acids and Polypeptides

One aspect of the present invention pertains to the use of isolated nucleic acid molecules that correspond to biomarker nucleic acids that encode a biomarker polypeptide or a portion of such a polypeptide. As used herein, the term "nucleic acid molecule" is intended to include DNA molecules (e.g., cDNA or genomic DNA) and RNA molecules (e.g., mRNA) and analogs of the DNA or RNA generated using nucleotide analogs. The nucleic acid molecule can be single-stranded or double-stranded, but preferably is double- stranded DNA.

An "isolated" nucleic acid molecule is one which is separated from other nucleic acid molecules which are present in the natural source of the nucleic acid molecule.

Preferably, an "isolated" nucleic acid molecule is free of sequences (preferably protein- encoding sequences) which naturally flank the nucleic acid (i.e., sequences located at the 5' and 3' ends of the nucleic acid) in the genomic DNA of the organism from which the nucleic acid is derived. For example, in various embodiments, the isolated nucleic acid molecule can contain less than about 5 kB, 4 kB, 3 kB, 2 kB, 1 kB, 0.5 kB or 0.1 kB of nucleotide sequences which naturally flank the nucleic acid molecule in genomic DNA of the cell from which the nucleic acid is derived. Moreover, an "isolated" nucleic acid molecule, such as a cDNA molecule, can be substantially free of other cellular material or culture medium when produced by recombinant techniques, or substantially free of chemical precursors or other chemicals when chemically synthesized.

A biomarker nucleic acid molecule of the present invention can be isolated using standard molecular biology techniques and the sequence information in the database records described herein. Using all or a portion of such nucleic acid sequences, nucleic acid molecules of the present invention can be isolated using standard hybridization and cloning techniques (e.g., as described in Sambrook et al, ed., Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1989).

A nucleic acid molecule of the present invention can be amplified using cDNA, mRNA, or genomic DNA as a template and appropriate oligonucleotide primers according to standard PCR amplification techniques. The nucleic acid molecules so amplified can be cloned into an appropriate vector and characterized by DNA sequence analysis.

Furthermore, oligonucleotides corresponding to all or a portion of a nucleic acid molecule of the present invention can be prepared by standard synthetic techniques, e.g., using an automated DNA synthesizer.

Moreover, a nucleic acid molecule of the present invention can comprise only a portion of a nucleic acid sequence, wherein the full length nucleic acid sequence comprises a marker of the present invention or which encodes a polypeptide corresponding to a marker of the present invention. Such nucleic acid molecules can be used, for example, as a probe or primer. The probe/primer typically is used as one or more substantially purified oligonucleotides. The oligonucleotide typically comprises a region of nucleotide sequence that hybridizes under stringent conditions to at least about 7, preferably about 15, more preferably about 25, 50, 75, 100, 125, 150, 175, 200, 250, 300, 350, or 400 or more consecutive nucleotides of a biomarker nucleic acid sequence. Probes based on the sequence of a biomarker nucleic acid molecule can be used to detect transcripts or genomic sequences corresponding to one or more markers of the present invention. The probe comprises a label group attached thereto, e.g., a radioisotope, a fluorescent compound, an enzyme, or an enzyme co-factor.

A biomarker nucleic acid molecules that differ, due to degeneracy of the genetic code, from the nucleotide sequence of nucleic acid molecules encoding a protein which corresponds to the biomarker, and thus encode the same protein, are also contemplated.

In addition, it will be appreciated by those skilled in the art that DNA sequence polymorphisms that lead to changes in the amino acid sequence can exist within a population {e.g., the human population). Such genetic polymorphisms can exist among individuals within a population due to natural allelic variation. An allele is one of a group of genes which occur alternatively at a given genetic locus. In addition, it will be appreciated that DNA polymorphisms that affect RNA expression levels can also exist that may affect the overall expression level of that gene {e.g., by affecting regulation or degradation). The term "allele," which is used interchangeably herein with "allelic variant," refers to alternative forms of a gene or portions thereof. Alleles occupy the same locus or position on homologous chromosomes. When a subject has two identical alleles of a gene, the subject is said to be homozygous for the gene or allele. When a subject has two different alleles of a gene, the subject is said to be heterozygous for the gene or allele. For example, biomarker alleles can differ from each other in a single nucleotide, or several nucleotides, and can include substitutions, deletions, and insertions of nucleotides. An allele of a gene can also be a form of a gene containing one or more mutations.

The term "allelic variant of a polymorphic region of gene" or "allelic variant", used interchangeably herein, refers to an alternative form of a gene having one of several possible nucleotide sequences found in that region of the gene in the population. As used herein, allelic variant is meant to encompass functional allelic variants, non-functional allelic variants, S Ps, mutations and polymorphisms.

The term "single nucleotide polymorphism" (S P) refers to a polymorphic site occupied by a single nucleotide, which is the site of variation between allelic sequences. The site is usually preceded by and followed by highly conserved sequences of the allele (e.g., sequences that vary in less than 1/100 or 1/1000 members of a population). A SNP usually arises due to substitution of one nucleotide for another at the polymorphic site. SNPs can also arise from a deletion of a nucleotide or an insertion of a nucleotide relative to a reference allele. Typically the polymorphic site is occupied by a base other than the reference base. For example, where the reference allele contains the base "T" (thymidine) at the polymorphic site, the altered allele can contain a "C" (cytidine), "G" (guanine), or "A" (adenine) at the polymorphic site. SNP's may occur in protein-coding nucleic acid sequences, in which case they may give rise to a defective or otherwise variant protein, or genetic disease. Such a SNP may alter the coding sequence of the gene and therefore specify another amino acid (a "missense" SNP) or a SNP may introduce a stop codon (a "nonsense" SNP). When a SNP does not alter the amino acid sequence of a protein, the SNP is called "silent." SNP's may also occur in noncoding regions of the nucleotide sequence. This may result in defective protein expression, e.g., as a result of alternative spicing, or it may have no effect on the function of the protein.

As used herein, the terms "gene" and "recombinant gene" refer to nucleic acid molecules comprising an open reading frame encoding a polypeptide corresponding to a marker of the present invention. Such natural allelic variations can typically result in 1-5% variance in the nucleotide sequence of a given gene. Alternative alleles can be identified by sequencing the gene of interest in a number of different individuals. This can be readily carried out by using hybridization probes to identify the same genetic locus in a variety of individuals. Any and all such nucleotide variations and resulting amino acid

polymorphisms or variations that are the result of natural allelic variation and that do not alter the functional activity are intended to be within the scope of the present invention.

In another embodiment, a biomarker nucleic acid molecule is at least 7, 15, 20, 25, 30, 40, 60, 80, 100, 150, 200, 250, 300, 350, 400, 450, 550, 650, 700, 800, 900, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2200, 2400, 2600, 2800, 3000, 3500, 4000, 4500, or more nucleotides in length and hybridizes under stringent conditions to a nucleic acid molecule corresponding to a marker of the present invention or to a nucleic acid molecule encoding a protein corresponding to a marker of the present invention. As used herein, the term "hybridizes under stringent conditions" is intended to describe conditions for hybridization and washing under which nucleotide sequences at least 60% (65%, 70%), 75%), 80%>, preferably 85%>) identical to each other typically remain hybridized to each other. Such stringent conditions are known to those skilled in the art and can be found in sections 6.3.1-6.3.6 of Current Protocols in Molecular Biology, John Wiley & Sons, N.Y. (1989). A preferred, non-limiting example of stringent hybridization conditions are hybridization in 6X sodium chloride/sodium citrate (SSC) at about 45°C, followed by one or more washes in 0.2X SSC, 0.1% SDS at 50-65°C.

In addition to naturally-occurring allelic variants of a nucleic acid molecule of the present invention that can exist in the population, the skilled artisan will further appreciate that sequence changes can be introduced by mutation thereby leading to changes in the amino acid sequence of the encoded protein, without altering the biological activity of the protein encoded thereby. For example, one can make nucleotide substitutions leading to amino acid substitutions at "non-essential" amino acid residues. A "non-essential" amino acid residue is a residue that can be altered from the wild-type sequence without altering the biological activity, whereas an "essential" amino acid residue is required for biological activity. For example, amino acid residues that are not conserved or only semi-conserved among homologs of various species may be non-essential for activity and thus would be likely targets for alteration. Alternatively, amino acid residues that are conserved among the homologs of various species {e.g., murine and human) may be essential for activity and thus would not be likely targets for alteration. Accordingly, another aspect of the present invention pertains to nucleic acid molecules encoding a polypeptide of the present invention that contain changes in amino acid residues that are not essential for activity. Such polypeptides differ in amino acid sequence from the naturally-occurring proteins which correspond to the markers of the present invention, yet retain biological activity. In one embodiment, a biomarker protein has an amino acid sequence that is at least about 40% identical, 50%, 60%, 70%, 75%, 80%, 83%, 85%, 87.5%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or identical to the amino acid sequence of a biomarker protein described herein.

An isolated nucleic acid molecule encoding a variant protein can be created by introducing one or more nucleotide substitutions, additions or deletions into the nucleotide sequence of nucleic acids of the present invention, such that one or more amino acid residue substitutions, additions, or deletions are introduced into the encoded protein. Mutations can be introduced by standard techniques, such as site-directed mutagenesis and PCR-mediated mutagenesis. Preferably, conservative amino acid substitutions are made at one or more predicted non-essential amino acid residues. A "conservative amino acid substitution" is one in which the amino acid residue is replaced with an amino acid residue having a similar side chain. Families of amino acid residues having similar side chains have been defined in the art. These families include amino acids with basic side chains (e.g., lysine, arginine, histidine), acidic side chains (e.g., aspartic acid, glutamic acid), uncharged polar side chains (e.g., glycine, asparagine, glutamine, serine, threonine, tyrosine, cysteine), non-polar side chains (e.g., alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine, tryptophan), beta-branched side chains (e.g., threonine, valine, isoleucine) and aromatic side chains (e.g., tyrosine, phenylalanine, tryptophan, histidine). Alternatively, mutations can be introduced randomly along all or part of the coding sequence, such as by saturation mutagenesis, and the resultant mutants can be screened for biological activity to identify mutants that retain activity. Following mutagenesis, the encoded protein can be expressed recombinantly and the activity of the protein can be determined.

In some embodiments, the present invention further contemplates the use of anti- biomarker antisense nucleic acid molecules, i.e., molecules which are complementary to a sense nucleic acid of the present invention, e.g., complementary to the coding strand of a double-stranded cDNA molecule corresponding to a marker of the present invention or complementary to an mRNA sequence corresponding to a marker of the present invention. Accordingly, an antisense nucleic acid molecule of the present invention can hydrogen bond to (i.e. anneal with) a sense nucleic acid of the present invention. The antisense nucleic acid can be complementary to an entire coding strand, or to only a portion thereof, e.g., all or part of the protein coding region (or open reading frame). An antisense nucleic acid molecule can also be antisense to all or part of a non-coding region of the coding strand of a nucleotide sequence encoding a polypeptide of the present invention. The non- coding regions ("5' and 3' untranslated regions") are the 5' and 3' sequences which flank the coding region and are not translated into amino acids.

An antisense oligonucleotide can be, for example, about 5, 10, 15, 20, 25, 30, 35, 40, 45, or 50 or more nucleotides in length. An antisense nucleic acid can be constructed using chemical synthesis and enzymatic ligation reactions using procedures known in the art. For example, an antisense nucleic acid (e.g., an antisense oligonucleotide) can be chemically synthesized using naturally occurring nucleotides or variously modified nucleotides designed to increase the biological stability of the molecules or to increase the physical stability of the duplex formed between the antisense and sense nucleic acids, e.g., phosphorothioate derivatives and acridine substituted nucleotides can be used. Examples of modified nucleotides which can be used to generate the antisense nucleic acid include 5- fluorouracil, 5-bromouracil, 5-chlorouracil, 5-iodouracil, hypoxanthine, xanthine, 4- acetylcytosine, 5-(carboxyhydroxylmethyl) uracil, 5-carboxymethylaminomethyl-2- thiouridine, 5-carboxymethylaminomethyluracil, dihydrouracil, beta-D-galactosylqueosine, inosine, N6-isopentenyladenine, 1-methylguanine, 1-methylinosine, 2,2-dimethylguanine,

2- methyladenine, 2-methylguanine, 3-methylcytosine, 5-methylcytosine, N6-adenine, 7- methylguanine, 5-methylaminomethyluracil, 5-methoxyaminomethyl-2-thiouracil, beta-D- mannosylqueosine, 5'-methoxycarboxymethyluracil, 5-methoxyuracil, 2-methylthio-N6- isopentenyladenine, uracil-5-oxyacetic acid (v), wybutoxosine, pseudouracil, queosine, 2- thiocytosine, 5-methyl-2-thiouracil, 2-thiouracil, 4-thiouracil, 5-methyluracil, uracil-5- oxyacetic acid methylester, uracil-5-oxyacetic acid (v), 5-methyl-2-thiouracil, 3-(3-amino-

3- N-2-carboxypropyl) uracil, (acp3)w, and 2,6-diaminopurine. Alternatively, the antisense nucleic acid can be produced biologically using an expression vector into which a nucleic acid has been sub-cloned in an antisense orientation (i.e., RNA transcribed from the inserted nucleic acid will be of an antisense orientation to a target nucleic acid of interest, described further in the following subsection).

The antisense nucleic acid molecules of the present invention are typically administered to a subject or generated in situ such that they hybridize with or bind to cellular mRNA and/or genomic DNA encoding a polypeptide corresponding to a selected marker of the present invention to thereby inhibit expression of the marker, e.g., by inhibiting transcription and/or translation. The hybridization can be by conventional nucleotide complementarity to form a stable duplex, or, for example, in the case of an antisense nucleic acid molecule which binds to DNA duplexes, through specific interactions in the major groove of the double helix. Examples of a route of administration of antisense nucleic acid molecules of the present invention includes direct injection at a tissue site or infusion of the antisense nucleic acid into a blood- or bone marrow-associated body fluid. Alternatively, antisense nucleic acid molecules can be modified to target selected cells and then administered systemically. For example, for systemic administration, antisense molecules can be modified such that they specifically bind to receptors or antigens expressed on a selected cell surface, e.g., by linking the antisense nucleic acid molecules to peptides or antibodies which bind to cell surface receptors or antigens. The antisense nucleic acid molecules can also be delivered to cells using the vectors described herein. To achieve sufficient intracellular concentrations of the antisense molecules, vector constructs in which the antisense nucleic acid molecule is placed under the control of a strong pol II or pol III promoter are preferred.

An antisense nucleic acid molecule of the present invention can be an a-anomeric nucleic acid molecule. An a-anomeric nucleic acid molecule forms specific double- stranded hybrids with complementary RNA in which, contrary to the usual a-units, the strands run parallel to each other (Gaultier et al. (1987) Nucleic Acids Res. 15:6625-6641). The antisense nucleic acid molecule can also comprise a 2'-o-methylribonucleotide (Inoue et al. (1987) Nucleic Acids Res. 15:6131-6148) or a chimeric RNA-DNA analogue (Inoue et al. (A9 1) FEBS Lett. 215:327-330).

The present invention also encompasses ribozymes. Ribozymes are catalytic RNA molecules with ribonuclease activity which are capable of cleaving a single-stranded nucleic acid, such as an mRNA, to which they have a complementary region. Thus, ribozymes {e.g., hammerhead ribozymes as described in Haselhoff and Gerlach (1988) Nature 334:585-591) can be used to catalytically cleave mRNA transcripts to thereby inhibit translation of the protein encoded by the mRNA. A ribozyme having specificity for a nucleic acid molecule encoding a polypeptide corresponding to a marker of the present invention can be designed based upon the nucleotide sequence of a cDNA corresponding to the marker. For example, a derivative of a Tetrahymena L-19 IVS RNA can be constructed in which the nucleotide sequence of the active site is complementary to the nucleotide sequence to be cleaved (see Cech et al. U.S. Patent No. 4,987,071; and Cech et al. U.S. Patent No. 5,116,742). Alternatively, an mRNA encoding a polypeptide of the present invention can be used to select a catalytic RNA having a specific ribonuclease activity from a pool of RNA molecules (see, e.g., Bartel and Szostak (1993) Science 261 : 1411-1418).

The present invention also encompasses nucleic acid molecules which form triple helical structures. For example, expression of a biomarker protein can be inhibited by targeting nucleotide sequences complementary to the regulatory region of the gene encoding the polypeptide {e.g., the promoter and/or enhancer) to form triple helical structures that prevent transcription of the gene in target cells. See generally Helene (1991) Anticancer Drug Des. 6(6):569-84; Helene (1992) Ann. NY. Acad. Sci. 660:27-36; and Maher ( 1992) Bioassays 14( 12) : 807- 15.

In various embodiments, the nucleic acid molecules of the present invention can be modified at the base moiety, sugar moiety or phosphate backbone to improve, e.g., the stability, hybridization, or solubility of the molecule. For example, the deoxyribose phosphate backbone of the nucleic acid molecules can be modified to generate peptide nucleic acid molecules (see Hyrup et al. (1996) Bioorganic & Medicinal Chemistry 4(1): 5- 23). As used herein, the terms "peptide nucleic acids" or "PNAs" refer to nucleic acid mimics, e.g., DNA mimics, in which the deoxyribose phosphate backbone is replaced by a pseudopeptide backbone and only the four natural nucleobases are retained. The neutral backbone of PNAs has been shown to allow for specific hybridization to DNA and RNA under conditions of low ionic strength. The synthesis of PNA oligomers can be performed using standard solid phase peptide synthesis protocols as described in Hyrup et al. (1996), supra; Perry-O'Keefe et al. (1996) Proc. Natl. Acad. Sci. USA 93 : 14670-675.

PNAs can be used in therapeutic and diagnostic applications. For example, PNAs can be used as antisense or antigene agents for sequence-specific modulation of gene expression by, e.g., inducing transcription or translation arrest or inhibiting replication. PNAs can also be used, e.g., in the analysis of single base pair mutations in a gene by, e.g., PNA directed PCR clamping; as artificial restriction enzymes when used in combination with other enzymes, e.g., SI nucleases (Hyrup (1996), supra; or as probes or primers for DNA sequence and hybridization (Hyrup (1996), supra; Perry-O'Keefe et al. (1996) Proc. Natl. Acad. Sci. USA 93 : 14670-14675). In another embodiment, PNAs can be modified, e.g., to enhance their stability or cellular uptake, by attaching lipophilic or other helper groups to PNA, by the formation of PNA-DNA chimeras, or by the use of liposomes or other techniques of drug delivery known in the art. For example, PNA-DNA chimeras can be generated which can combine the advantageous properties of PNA and DNA. Such chimeras allow DNA recognition enzymes, e.g., RNASE H and DNA polymerases, to interact with the DNA portion while the PNA portion would provide high binding affinity and specificity. PNA-DNA chimeras can be linked using linkers of appropriate lengths selected in terms of base stacking, number of bonds between the nucleobases, and orientation (Hyrup (1996), supra). The synthesis of PNA-DNA chimeras can be performed as described in Hyrup (1996), supra, and Finn et al. (1996) Nucleic Acids Res. 24(17):3357-3363. For example, a DNA chain can be synthesized on a solid support using standard phosphoramidite coupling chemistry and modified nucleoside analogs. Compounds such as 5'-(4-methoxytrityl)amino-5'-deoxy- thymidine phosphoramidite can be used as a link between the PNA and the 5' end of DNA (Mag et al. (1989) Nucleic Acids Res. 17:5973-5988). PNA monomers are then coupled in a step-wise manner to produce a chimeric molecule with a 5' PNA segment and a 3' DNA segment (Finn et al. (1996) Nucleic Acids Res. 24:3357-3363). Alternatively, chimeric molecules can be synthesized with a 5' DNA segment and a 3' PNA segment (Peterser et al. (1975) Bioorganic Med. Chem. Lett. 5: 1119-11124).

In other embodiments, the oligonucleotide can include other appended groups such as peptides {e.g., for targeting host cell receptors in vivo), or agents facilitating transport across the cell membrane (see, e.g., Letsinger et al. (1989) Proc. Natl. Acad. Sci. USA 86:6553-6556; Lemaitre et al. (1987) Proc. Natl. Acad. Sci. USA 84:648-652; PCT

Publication No. WO 88/09810) or the blood-brain barrier (see, e.g., PCT Publication No. WO 89/10134). In addition, oligonucleotides can be modified with hybridization-triggered cleavage agents (see, e.g., Krol et al. (1988) Bio/Techniques 6:958-976) or intercalating agents (see, e.g., Zon (1988) Pharm. Res. 5:539-549). To this end, the oligonucleotide can be conjugated to another molecule, e.g., a peptide, hybridization triggered cross-linking agent, transport agent, hybridization-triggered cleavage agent, etc.

Another aspect of the present invention pertains to the use of biomarker proteins and biologically active portions thereof. In one embodiment, the native polypeptide

corresponding to a marker can be isolated from cells or tissue sources by an appropriate purification scheme using standard protein purification techniques. In another embodiment, polypeptides corresponding to a marker of the present invention are produced by

recombinant DNA techniques. Alternative to recombinant expression, a polypeptide corresponding to a marker of the present invention can be synthesized chemically using standard peptide synthesis techniques.

An "isolated" or "purified" protein or biologically active portion thereof is substantially free of cellular material or other contaminating proteins from the cell or tissue source from which the protein is derived, or substantially free of chemical precursors or other chemicals when chemically synthesized. The language "substantially free of cellular material" includes preparations of protein in which the protein is separated from cellular components of the cells from which it is isolated or recombinantly produced. Thus, protein that is substantially free of cellular material includes preparations of protein having less than about 30%, 20%, 10%, or 5% (by dry weight) of heterologous protein (also referred to herein as a "contaminating protein"). When the protein or biologically active portion thereof is recombinantly produced, it is also preferably substantially free of culture medium, i.e., culture medium represents less than about 20%, 10%, or 5% of the volume of the protein preparation. When the protein is produced by chemical synthesis, it is preferably substantially free of chemical precursors or other chemicals, i.e., it is separated from chemical precursors or other chemicals which are involved in the synthesis of the protein. Accordingly such preparations of the protein have less than about 30%, 20%, 10%, 5% (by dry weight) of chemical precursors or compounds other than the polypeptide of interest.

Biologically active portions of a biomarker polypeptide include polypeptides comprising amino acid sequences sufficiently identical to or derived from a biomarker protein amino acid sequence described herein, but which includes fewer amino acids than the full length protein, and exhibit at least one activity of the corresponding full-length protein. Typically, biologically active portions comprise a domain or motif with at least one activity of the corresponding protein. A biologically active portion of a protein of the present invention can be a polypeptide which is, for example, 10, 25, 50, 100 or more amino acids in length. Moreover, other biologically active portions, in which other regions of the protein are deleted, can be prepared by recombinant techniques and evaluated for one or more of the functional activities of the native form of a polypeptide of the present invention. Preferred polypeptides have an amino acid sequence of a biomarker protein encoded by a nucleic acid molecule described herein. Other useful proteins are substantially identical (e.g., at least about 40%, preferably 50%, 60%, 70%, 75%, 80%, 83%, 85%, 88%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) to one of these sequences and retain the functional activity of the protein of the corresponding naturally-occurring protein yet differ in amino acid sequence due to natural allelic variation or mutagenesis.

To determine the percent identity of two amino acid sequences or of two nucleic acids, the sequences are aligned for optimal comparison purposes (e.g., gaps can be introduced in the sequence of a first amino acid or nucleic acid sequence for optimal alignment with a second amino or nucleic acid sequence). The amino acid residues or nucleotides at corresponding amino acid positions or nucleotide positions are then compared. When a position in the first sequence is occupied by the same amino acid residue or nucleotide as the corresponding position in the second sequence, then the molecules are identical at that position. The percent identity between the two sequences is a function of the number of identical positions shared by the sequences (i.e., % identity = # of identical positions/total # of positions (e.g., overlapping positions) xlOO). In one embodiment the two sequences are the same length.

The determination of percent identity between two sequences can be accomplished using a mathematical algorithm. A preferred, non-limiting example of a mathematical algorithm utilized for the comparison of two sequences is the algorithm of Karlin and Altschul (1990) Proc. Natl. Acad. Sci. USA 87:2264-2268, modified as in Karlin and Altschul (1993) Proc. Natl. Acad. Sci. USA 90:5873-5877. Such an algorithm is incorporated into the BLAST and XBLAST programs of Altschul, et al. (1990) J. Mol. Biol. 215:403-410. BLAST nucleotide searches can be performed with the NBLAST program, score = 100, wordlength = 12 to obtain nucleotide sequences homologous to a nucleic acid molecules of the present invention. BLAST protein searches can be performed with the XBLAST program, score = 50, wordlength = 3 to obtain amino acid sequences homologous to a protein molecules of the present invention. To obtain gapped alignments for comparison purposes, Gapped BLAST can be utilized as described in Altschul et al. (1997) Nucleic Acids Res. 25:3389-3402. Alternatively, PSI-Blast can be used to perform an iterated search which detects distant relationships between molecules. When utilizing BLAST, Gapped BLAST, and PSI-Blast programs, the default parameters of the respective programs (e.g., XBLAST and NBLAST) can be used. See http://www.ncbi.nlm.nih.gov. Another preferred, non-limiting example of a mathematical algorithm utilized for the comparison of sequences is the algorithm of Myers and Miller, (1988) Comput Appl Biosci, 4: 11-7. Such an algorithm is incorporated into the ALIGN program (version 2.0) which is part of the GCG sequence alignment software package. When utilizing the ALIGN program for comparing amino acid sequences, a PAM120 weight residue table, a gap length penalty of 12, and a gap penalty of 4 can be used. Yet another useful algorithm for identifying regions of local sequence similarity and alignment is the FASTA algorithm as described in Pearson and Lipman (1988) Proc. Natl. Acad. Sci. USA 85:2444-2448. When using the FASTA algorithm for comparing nucleotide or amino acid sequences, a PAM120 weight residue table can, for example, be used with a &-tuple value of 2.

The percent identity between two sequences can be determined using techniques similar to those described above, with or without allowing gaps. In calculating percent identity, only exact matches are counted.

The present invention also provides chimeric or fusion proteins corresponding to a biomarker protein. As used herein, a "chimeric protein" or "fusion protein" comprises all or part (preferably a biologically active part) of a polypeptide corresponding to a marker of the present invention operably linked to a heterologous polypeptide (i.e., a polypeptide other than the polypeptide corresponding to the marker). Within the fusion protein, the term "operably linked" is intended to indicate that the polypeptide of the present invention and the heterologous polypeptide are fused in-frame to each other. The heterologous polypeptide can be fused to the amino-terminus or the carboxyl-terminus of the polypeptide of the present invention.

One useful fusion protein is a GST fusion protein in which a polypeptide corresponding to a marker of the present invention is fused to the carboxyl terminus of GST sequences. Such fusion proteins can facilitate the purification of a recombinant polypeptide of the present invention.

In another embodiment, the fusion protein contains a heterologous signal sequence, immunoglobulin fusion protein, toxin, or other useful protein sequence. Chimeric and fusion proteins of the present invention can be produced by standard recombinant DNA techniques. In another embodiment, the fusion gene can be synthesized by conventional techniques including automated DNA synthesizers. Alternatively, PCR amplification of gene fragments can be carried out using anchor primers which give rise to complementary overhangs between two consecutive gene fragments which can subsequently be annealed and re-amplified to generate a chimeric gene sequence (see, e.g., Ausubel et al, supra). Moreover, many expression vectors are commercially available that already encode a fusion moiety {e.g., a GST polypeptide). A nucleic acid encoding a polypeptide of the present invention can be cloned into such an expression vector such that the fusion moiety is linked in-frame to the polypeptide of the present invention.

A signal sequence can be used to facilitate secretion and isolation of the secreted protein or other proteins of interest. Signal sequences are typically characterized by a core of hydrophobic amino acids which are generally cleaved from the mature protein during secretion in one or more cleavage events. Such signal peptides contain processing sites that allow cleavage of the signal sequence from the mature proteins as they pass through the secretory pathway. Thus, the present invention pertains to the described polypeptides having a signal sequence, as well as to polypeptides from which the signal sequence has been proteolytically cleaved {i.e., the cleavage products). In one embodiment, a nucleic acid sequence encoding a signal sequence can be operably linked in an expression vector to a protein of interest, such as a protein which is ordinarily not secreted or is otherwise difficult to isolate. The signal sequence directs secretion of the protein, such as from a eukaryotic host into which the expression vector is transformed, and the signal sequence is subsequently or concurrently cleaved. The protein can then be readily purified from the extracellular medium by art recognized methods. Alternatively, the signal sequence can be linked to the protein of interest using a sequence which facilitates purification, such as with a GST domain.

The present invention also pertains to variants of the biomarker polypeptides described herein. Such variants have an altered amino acid sequence which can function as either agonists (mimetics) or as antagonists. Variants can be generated by mutagenesis, e.g., discrete point mutation or truncation. An agonist can retain substantially the same, or a subset, of the biological activities of the naturally occurring form of the protein. An antagonist of a protein can inhibit one or more of the activities of the naturally occurring form of the protein by, for example, competitively binding to a downstream or upstream member of a cellular signaling cascade which includes the protein of interest. Thus, specific biological effects can be elicited by treatment with a variant of limited function. Treatment of a subject with a variant having a subset of the biological activities of the naturally occurring form of the protein can have fewer side effects in a subject relative to treatment with the naturally occurring form of the protein. Variants of a biomarker protein which function as either agonists (mimetics) or as antagonists can be identified by screening combinatorial libraries of mutants, e.g., truncation mutants, of the protein of the present invention for agonist or antagonist activity. In one embodiment, a variegated library of variants is generated by combinatorial mutagenesis at the nucleic acid level and is encoded by a variegated gene library. A variegated library of variants can be produced by, for example, enzymatically ligating a mixture of synthetic oligonucleotides into gene sequences such that a degenerate set of potential protein sequences is expressible as individual polypeptides, or alternatively, as a set of larger fusion proteins (e.g., for phage display). There are a variety of methods which can be used to produce libraries of potential variants of the polypeptides of the present invention from a degenerate oligonucleotide sequence. Methods for synthesizing degenerate oligonucleotides are known in the art (see, e.g., Narang (1983) Tetrahedron 39:3; Itakura et al. (1984) Annu. Rev. Biochem. 53 :323; Itakura et al. (1984) Science 198: 1056; Ike et al. (1983) Nucleic Acid Res. 11 :477).

In addition, libraries of fragments of the coding sequence of a polypeptide corresponding to a marker of the present invention can be used to generate a variegated population of polypeptides for screening and subsequent selection of variants. For example, a library of coding sequence fragments can be generated by treating a double stranded PCR fragment of the coding sequence of interest with a nuclease under conditions wherein nicking occurs only about once per molecule, denaturing the double stranded

DNA, renaturing the DNA to form double stranded DNA which can include sense/antisense pairs from different nicked products, removing single stranded portions from reformed duplexes by treatment with SI nuclease, and ligating the resulting fragment library into an expression vector. By this method, an expression library can be derived which encodes amino terminal and internal fragments of various sizes of the protein of interest.

Several techniques are known in the art for screening gene products of

combinatorial libraries made by point mutations or truncation, and for screening cDNA libraries for gene products having a selected property. The most widely used techniques, which are amenable to high throughput analysis, for screening large gene libraries typically include cloning the gene library into replicable expression vectors, transforming appropriate cells with the resulting library of vectors, and expressing the combinatorial genes under conditions in which detection of a desired activity facilitates isolation of the vector encoding the gene whose product was detected. Recursive ensemble mutagenesis (REM), a technique which enhances the frequency of functional mutants in the libraries, can be used in combination with the screening assays to identify variants of a protein of the present invention (Arkin and Yourvan (1992) Proc. Natl. Acad. Sci. USA 59:7811-7815; Delgrave et al. 91993) Protein Engineering 6(3): 327- 331).

The production and use of biomarker nucleic acid and/or biomarker polypeptide molecules described herein can be facilitated by using standard recombinant techniques. In some embodiments, such techniques use vectors, preferably expression vectors, containing a nucleic acid encoding a biomarker polypeptide or a portion of such a polypeptide. As used herein, the term "vector" refers to a nucleic acid molecule capable of transporting another nucleic acid to which it has been linked. One type of vector is a "plasmid", which refers to a circular double stranded DNA loop into which additional DNA segments can be ligated. Another type of vector is a viral vector, wherein additional DNA segments can be ligated into the viral genome. Certain vectors are capable of autonomous replication in a host cell into which they are introduced {e.g., bacterial vectors having a bacterial origin of replication and episomal mammalian vectors). Other vectors {e.g., non-episomal mammalian vectors) are integrated into the genome of a host cell upon introduction into the host cell, and thereby are replicated along with the host genome. Moreover, certain vectors, namely expression vectors, are capable of directing the expression of genes to which they are operably linked. In general, expression vectors of utility in recombinant DNA techniques are often in the form of plasmids (vectors). However, the present invention is intended to include such other forms of expression vectors, such as viral vectors {e.g., replication defective retroviruses, adenoviruses and adeno-associated viruses), which serve equivalent functions.

The recombinant expression vectors of the present invention comprise a nucleic acid of the present invention in a form suitable for expression of the nucleic acid in a host cell. This means that the recombinant expression vectors include one or more regulatory sequences, selected on the basis of the host cells to be used for expression, which is operably linked to the nucleic acid sequence to be expressed. Within a recombinant expression vector, "operably linked" is intended to mean that the nucleotide sequence of interest is linked to the regulatory sequence(s) in a manner which allows for expression of the nucleotide sequence {e.g., in an in vitro transcription/translation system or in a host cell when the vector is introduced into the host cell). The term "regulatory sequence" is intended to include promoters, enhancers and other expression control elements {e.g., polyadenylation signals). Such regulatory sequences are described, for example, in

Goeddel, Methods in Enzymology: Gene Expression Technology vol.185, Academic Press, San Diego, CA (1991). Regulatory sequences include those which direct constitutive expression of a nucleotide sequence in many types of host cell and those which direct expression of the nucleotide sequence only in certain host cells {e.g., tissue-specific regulatory sequences). It will be appreciated by those skilled in the art that the design of the expression vector can depend on such factors as the choice of the host cell to be transformed, the level of expression of protein desired, and the like. The expression vectors of the present invention can be introduced into host cells to thereby produce proteins or peptides, including fusion proteins or peptides, encoded by nucleic acids as described herein.

The recombinant expression vectors for use in the present invention can be designed for expression of a polypeptide corresponding to a marker of the present invention in prokaryotic {e.g., E. coli) or eukaryotic cells {e.g., insect cells {using baculovirus expression vectors}, yeast cells or mammalian cells). Suitable host cells are discussed further in Goeddel, supra. Alternatively, the recombinant expression vector can be transcribed and translated in vitro, for example using T7 promoter regulatory sequences and T7 polymerase.

Expression of proteins in prokaryotes is most often carried out in E. coli with vectors containing constitutive or inducible promoters directing the expression of either fusion or non-fusion proteins. Fusion vectors add a number of amino acids to a protein encoded therein, usually to the amino terminus of the recombinant protein. Such fusion vectors typically serve three purposes: 1) to increase expression of recombinant protein; 2) to increase the solubility of the recombinant protein; and 3) to aid in the purification of the recombinant protein by acting as a ligand in affinity purification. Often, in fusion expression vectors, a proteolytic cleavage site is introduced at the junction of the fusion moiety and the recombinant protein to enable separation of the recombinant protein from the fusion moiety subsequent to purification of the fusion protein. Such enzymes, and their cognate recognition sequences, include Factor Xa, thrombin and enterokinase. Typical fusion expression vectors include pGEX (Pharmacia Biotech Inc; Smith and Johnson, 1988, Gene 67:31-40), pMAL (New England Biolabs, Beverly, MA) and pRIT5 (Pharmacia, Piscataway, NJ) which fuse glutathione S-transferase (GST), maltose E binding protein, or protein A, respectively, to the target recombinant protein. Examples of suitable inducible non-fusion E. coli expression vectors include pTrc (Amann et a/. (1988) Gene 69:301-315) and pET 1 Id (Studier et al., p. 60-89, In Gene Expression Technology: Methods in Enzymology vol.185, Academic Press, San Diego, CA, 1991). Target biomarker nucleic acid expression from the pTrc vector relies on host RNA polymerase transcription from a hybrid trp-lac fusion promoter. Target biomarker nucleic acid expression from the pET l id vector relies on transcription from a T7 gnlO-lac fusion promoter mediated by a co-expressed viral RNA polymerase (T7 gnl). This viral polymerase is supplied by host strains BL21 (DE3) or HMS174(DE3) from a resident prophage harboring a T7 gnl gene under the transcriptional control of the lacUV 5 promoter.

One strategy to maximize recombinant protein expression in E. coli is to express the protein in a host bacterium with an impaired capacity to proteolytically cleave the recombinant protein (Gottesman, p. 119-128, In Gene Expression Technology: Methods in Enzymology vol. 185, Academic Press, San Diego, CA, 1990. Another strategy is to alter the nucleic acid sequence of the nucleic acid to be inserted into an expression vector so that the individual codons for each amino acid are those preferentially utilized in E. coli (Wada et al, (1992) Nucleic Acids Res. 20:2111-2118). Such alteration of nucleic acid sequences of the present invention can be carried out by standard DNA synthesis techniques.

In another embodiment, the expression vector is a yeast expression vector.

Examples of vectors for expression in yeast S. cerevisiae include pYepSecl (Baldari et al. (1987) EMBO J. 6:229-234), pMFa (Kurjan and Herskowitz (1982) Cell 30:933-943), pJRY88 (Schultz et al. (1987) Gene 54: 113-123), pYES2 (Invitrogen Corporation, San Diego, CA), and pPicZ (Invitrogen Corp, San Diego, CA).

Alternatively, the expression vector is a baculovirus expression vector. Baculovirus vectors available for expression of proteins in cultured insect cells {e.g., Sf 9 cells) include the pAc series (Smith et al. (1983) o/. Cell Biol. 3 :2156-2165) and the pVL series (Lucklow and Summers (1989) Virology 170:31-39).

In yet another embodiment, a nucleic acid of the present invention is expressed in mammalian cells using a mammalian expression vector. Examples of mammalian expression vectors include pCDM8 (Seed (1987) Nature 329:840) and pMT2PC (Kaufman et al. (1987) EMBO J. 6: 187-195). When used in mammalian cells, the expression vector's control functions are often provided by viral regulatory elements. For example, commonly used promoters are derived from polyoma, Adenovirus 2, cytomegalovirus and Simian Virus 40. For other suitable expression systems for both prokaryotic and eukaryotic cells see chapters 16 and 17 of Sambrook et al, supra.

In another embodiment, the recombinant mammalian expression vector is capable of directing expression of the nucleic acid preferentially in a particular cell type {e.g., tissue- specific regulatory elements are used to express the nucleic acid). Tissue-specific regulatory elements are known in the art. Non-limiting examples of suitable tissue-specific promoters include the albumin promoter (liver-specific; Pinkert et al. (1987) Genes Dev. 1 :268-277), lymphoid-specific promoters (Calame and Eaton (1988) Adv. Immunol. 43 :235- 275), in particular promoters of T cell receptors (Winoto and Baltimore (1989) EMBO J. 8:729-733) and immunoglobulins (Banerji et al. (1983) Cell 33 :729-740; Queen and Baltimore (1983) Cell 33 :741-748), neuron-specific promoters {e.g., the neurofilament promoter; Byrne and Ruddle (1989) Proc. Natl. Acad. Sci. USA 86:5473-5477), pancreas- specific promoters (Edlund et al. (1985) Science 230:912-916), and mammary gland- specific promoters {e.g., milk whey promoter; U.S. Patent No. 4,873,316 and European Application Publication No. 264,166). Developmentally-regulated promoters are also encompassed, for example the murine hox promoters (Kessel and Gruss (1990) Science 249:374-379) and the a-fetoprotein promoter (Camper and Tilghman (1989) Genes Dev. 3 :537-546).

The present invention further provides a recombinant expression vector comprising a DNA molecule cloned into the expression vector in an antisense orientation. That is, the DNA molecule is operably linked to a regulatory sequence in a manner which allows for expression (by transcription of the DNA molecule) of an RNA molecule which is antisense to the mRNA encoding a polypeptide of the present invention. Regulatory sequences operably linked to a nucleic acid cloned in the antisense orientation can be chosen which direct the continuous expression of the antisense RNA molecule in a variety of cell types, for instance viral promoters and/or enhancers, or regulatory sequences can be chosen which direct constitutive, tissue-specific or cell type specific expression of antisense RNA. The antisense expression vector can be in the form of a recombinant plasmid, phagemid, or attenuated virus in which antisense nucleic acids are produced under the control of a high efficiency regulatory region, the activity of which can be determined by the cell type into which the vector is introduced. For a discussion of the regulation of gene expression using antisense genes (see Weintraub et al. (1986) Trends in Genetics, Vol. 1(1)). Another aspect of the present invention pertains to host cells into which a recombinant expression vector of the present invention has been introduced. The terms "host cell" and "recombinant host cell" are used interchangeably herein. It is understood that such terms refer not only to the particular subject cell but to the progeny or potential progeny of such a cell. Because certain modifications may occur in succeeding generations due to either mutation or environmental influences, such progeny may not, in fact, be identical to the parent cell, but are still included within the scope of the term as used herein.

A host cell can be any prokaryotic (e.g., E. coli) or eukaryotic cell {e.g., insect cells, yeast or mammalian cells).

Vector DNA can be introduced into prokaryotic or eukaryotic cells via conventional transformation or transfection techniques. As used herein, the terms "transformation" and "transfection" are intended to refer to a variety of art-recognized techniques for introducing foreign nucleic acid into a host cell, including calcium phosphate or calcium chloride co- precipitation, DEAE-dextran-mediated transfection, lipofection, or electroporation.

Suitable methods for transforming or transfecting host cells can be found in Sambrook, et al. {supra), and other laboratory manuals.

For stable transfection of mammalian cells, it is known that, depending upon the expression vector and transfection technique used, only a small fraction of cells may integrate the foreign DNA into their genome. In order to identify and select these integrants, a gene that encodes a selectable marker {e.g., for resistance to antibiotics) is generally introduced into the host cells along with the gene of interest. Preferred selectable markers include those which confer resistance to drugs, such as G418, hygromycin and methotrexate. Cells stably transfected with the introduced nucleic acid can be identified by drug selection {e.g., cells that have incorporated the selectable marker gene will survive, while the other cells die).

V. Analyzing Biomarker Nucleic Acids and Polypeptides

Biomarker nucleic acids and/or biomarker polypeptides can be analyzed according to the methods described herein and techniques known to the skilled artisan to identify such genetic or expression alterations useful for the present invention including, but not limited to, 1) an alteration in the level of a biomarker transcript or polypeptide, 2) a deletion or addition of one or more nucleotides from a biomarker gene, 4) a substitution of one or more nucleotides of a biomarker gene, 5) aberrant modification of a biomarker gene, such as an expression regulatory region, and the like.

a. Methods for Detection of Copy Number

Methods of evaluating the copy number of a biomarker nucleic acid are well known to those of skill in the art. The presence or absence of chromosomal gain or loss can be evaluated simply by a determination of copy number of the regions or markers identified herein.

In one embodiment, a biological sample is tested for the presence of copy number changes in genomic loci containing the genomic marker. A copy number of at least 3, 4, 5, 6, 7, 8, 9, or 10 is predictive of poorer outcome of anti-immune checkpoint treatment.

Methods of evaluating the copy number of a biomarker locus include, but are not limited to, hybridization-based assays. Hybridization-based assays include, but are not limited to, traditional "direct probe" methods, such as Southern blots, in situ hybridization {e.g., FISH and FISH plus SKY) methods, and "comparative probe" methods, such as comparative genomic hybridization (CGH), e.g., cDNA-based or oligonucleotide-based CGH. The methods can be used in a wide variety of formats including, but not limited to, substrate {e.g. membrane or glass) bound methods or array-based approaches.

In one embodiment, evaluating the biomarker gene copy number in a sample involves a Southern Blot. In a Southern Blot, the genomic DNA (typically fragmented and separated on an electrophoretic gel) is hybridized to a probe specific for the target region. Comparison of the intensity of the hybridization signal from the probe for the target region with control probe signal from analysis of normal genomic DNA {e.g., a non-amplified portion of the same or related cell, tissue, organ, etc) provides an estimate of the relative copy number of the target nucleic acid. Alternatively, a Northern blot may be utilized for evaluating the copy number of encoding nucleic acid in a sample. In a Northern blot, mRNA is hybridized to a probe specific for the target region. Comparison of the intensity of the hybridization signal from the probe for the target region with control probe signal from analysis of normal RNA {e.g., a non-amplified portion of the same or related cell, tissue, organ, etc) provides an estimate of the relative copy number of the target nucleic acid. Alternatively, other methods well known in the art to detect RNA can be used, such that higher or lower expression relative to an appropriate control {e.g., a non-amplified portion of the same or related cell tissue, organ, etc.) provides an estimate of the relative copy number of the target nucleic acid. An alternative means for determining genomic copy number is in situ hybridization {e.g., Angerer (1987) et/z. Enzymol 152: 649). Generally, in situ hybridization comprises the following steps: (1) fixation of tissue or biological structure to be analyzed; (2) prehybridization treatment of the biological structure to increase accessibility of target DNA, and to reduce nonspecific binding; (3) hybridization of the mixture of nucleic acids to the nucleic acid in the biological structure or tissue; (4) post-hybridization washes to remove nucleic acid fragments not bound in the hybridization and (5) detection of the hybridized nucleic acid fragments. The reagent used in each of these steps and the conditions for use vary depending on the particular application. In a typical in situ hybridization assay, cells are fixed to a solid support, typically a glass slide. If a nucleic acid is to be probed, the cells are typically denatured with heat or alkali. The cells are then contacted with a hybridization solution at a moderate temperature to permit annealing of labeled probes specific to the nucleic acid sequence encoding the protein. The targets {e.g., cells) are then typically washed at a predetermined stringency or at an increasing stringency until an appropriate signal to noise ratio is obtained. The probes are typically labeled, e.g., with radioisotopes or fluorescent reporters. In one embodiment, probes are sufficiently long so as to specifically hybridize with the target nucleic acid(s) under stringent conditions. Probes generally range in length from about 200 bases to about 1000 bases. In some applications it is necessary to block the hybridization capacity of repetitive sequences. Thus, in some embodiments, tRNA, human genomic DNA, or Cot-I DNA is used to block non-specific hybridization.

An alternative means for determining genomic copy number is comparative genomic hybridization. In general, genomic DNA is isolated from normal reference cells, as well as from test cells {e.g., tumor cells) and amplified, if necessary. The two nucleic acids are differentially labeled and then hybridized in situ to metaphase chromosomes of a reference cell. The repetitive sequences in both the reference and test DNAs are either removed or their hybridization capacity is reduced by some means, for example by prehybridization with appropriate blocking nucleic acids and/or including such blocking nucleic acid sequences for said repetitive sequences during said hybridization. The bound, labeled DNA sequences are then rendered in a visualizable form, if necessary.

Chromosomal regions in the test cells which are at increased or decreased copy number can be identified by detecting regions where the ratio of signal from the two DNAs is altered. For example, those regions that have decreased in copy number in the test cells will show relatively lower signal from the test DNA than the reference compared to other regions of the genome. Regions that have been increased in copy number in the test cells will show relatively higher signal from the test DNA. Where there are chromosomal deletions or multiplications, differences in the ratio of the signals from the two labels will be detected and the ratio will provide a measure of the copy number. In another embodiment of CGH, array CGH (aCGH), the immobilized chromosome element is replaced with a collection of solid support bound target nucleic acids on an array, allowing for a large or complete percentage of the genome to be represented in the collection of solid support bound targets. Target nucleic acids may comprise cDNAs, genomic DNAs, oligonucleotides (e.g., to detect single nucleotide polymorphisms) and the like. Array -based CGH may also be performed with single-color labeling (as opposed to labeling the control and the possible tumor sample with two different dyes and mixing them prior to hybridization, which will yield a ratio due to competitive hybridization of probes on the arrays). In single color CGH, the control is labeled and hybridized to one array and absolute signals are read, and the possible tumor sample is labeled and hybridized to a second array (with identical content) and absolute signals are read. Copy number difference is calculated based on absolute signals from the two arrays. Methods of preparing immobilized chromosomes or arrays and performing comparative genomic hybridization are well known in the art (see, e.g., U.S. Pat. Nos: 6,335,167; 6,197,501; 5,830,645; and 5,665,549 and Albertson (1984) EMBO J. 3 : 1227-1234; Pinkel (1988) roc. Natl. Acad. Sci. USA 85: 9138-9142; EPO

Pub. No. 430,402; Methods in Molecular Biology, Vol. 33 : In situ Hybridization Protocols, Choo, ed., Humana Press, Totowa, N.J. (1994), etc.) In another embodiment, the hybridization protocol of Pinkel, et al. (1998) Nature Genetics 20: 207-211, or of

Kallioniemi (1992) Proc. Natl Acad Sci USA 89:5321-5325 (1992) is used.

In still another embodiment, amplification-based assays can be used to measure copy number. In such amplification-based assays, the nucleic acid sequences act as a template in an amplification reaction {e.g., Polymerase Chain Reaction (PCR). In a quantitative amplification, the amount of amplification product will be proportional to the amount of template in the original sample. Comparison to appropriate controls, e.g. healthy tissue, provides a measure of the copy number.

Methods of "quantitative" amplification are well known to those of skill in the art. For example, quantitative PCR involves simultaneously co-amplifying a known quantity of a control sequence using the same primers. This provides an internal standard that may be used to calibrate the PCR reaction. Detailed protocols for quantitative PCR are provided in Innis, et al. (1990) PCR Protocols, A Guide to Methods and Applications, Academic Press, Inc. N.Y.). Measurement of DNA copy number at microsatellite loci using quantitative PCR analysis is described in Ginzonger, et al. (2000) Cancer Research 60:5405-5409. The known nucleic acid sequence for the genes is sufficient to enable one of skill in the art to routinely select primers to amplify any portion of the gene. Fluorogenic quantitative PCR may also be used in the methods of the present invention. In fluorogenic quantitative PCR, quantitation is based on amount of fluorescence signals, e.g., TaqMan and SYBR green.

Other suitable amplification methods include, but are not limited to, ligase chain reaction (LCR) (see Wu and Wallace (1989) Genomics 4: 560, Landegren, et al. (1988) Science 241 : 1077, and Barringer et al. (1990) Gene 89: 117), transcription amplification (Kwoh, et al. (1989) Proc. Natl. Acad. Sci. USA 86: 1173), self-sustained sequence replication (Guatelli, et al. (1990) Proc. Nat. Acad. Sci. USA 87: 1874), dot PCR, and linker adapter PCR, etc.

Loss of heterozygosity (LOH) and major copy proportion (MCP) mapping (Wang,

Z.C., et al. (2004) Cancer Res 64(1):64-71; Seymour, A. B., et al. (1994) Cancer Res 54, 2761-4; Hahn, S. A., et al. (1995) Cancer Res 55, 4670-5; Kimura, M., et al. (1996) Genes Chromosomes Cancer 17, 88-93; Li et al., (2008) MBC Bioinform. 9, 204-219) may also be used to identify regions of amplification or deletion.

b. Methods for Detection of Biomarker Nucleic Acid Expression

Biomarker expression may be assessed by any of a wide variety of well known methods for detecting expression of a transcribed molecule or protein. Non-limiting examples of such methods include immunological methods for detection of secreted, cell- surface, cytoplasmic, or nuclear proteins, protein purification methods, protein function or activity assays, nucleic acid hybridization methods, nucleic acid reverse transcription methods, and nucleic acid amplification methods.

In preferred embodiments, activity of a particular gene is characterized by a measure of gene transcript {e.g. mRNA), by a measure of the quantity of translated protein, or by a measure of gene product activity. Marker expression can be monitored in a variety of ways, including by detecting mRNA levels, protein levels, or protein activity, any of which can be measured using standard techniques. Detection can involve quantification of the level of gene expression {e.g., genomic DNA, cDNA, mRNA, protein, or enzyme activity), or, alternatively, can be a qualitative assessment of the level of gene expression, in particular in comparison with a control level. The type of level being detected will be clear from the context.

In another embodiment, detecting or determining expression levels of a biomarker and functionally similar homologs thereof, including a fragment or genetic alteration thereof (e.g., in regulatory or promoter regions thereof) comprises detecting or determining RNA levels for the marker of interest. In one embodiment, one or more cells from the subject to be tested are obtained and RNA is isolated from the cells. In a preferred embodiment, a sample of breast tissue cells is obtained from the subject.

In one embodiment, RNA is obtained from a single cell. For example, a cell can be isolated from a tissue sample by laser capture microdissection (LCM). Using this technique, a cell can be isolated from a tissue section, including a stained tissue section, thereby assuring that the desired cell is isolated (see, e.g., Bonner et al. (1997) Science 278: 1481; Emmert-Buck et al. (1996) Science 274:998; Fend et al. (1999) Am. J. Path. 154:61 and Murakami et al. (2000) Kidney Int. 58: 1346). For example, Murakami et al., supra, describe isolation of a cell from a previously immunostained tissue section.

It is also possible to obtain cells from a subject and culture the cells in vitro, such as to obtain a larger population of cells from which RNA can be extracted. Methods for establishing cultures of non-transformed cells, i.e., primary cell cultures, are known in the art.

When isolating RNA from tissue samples or cells from individuals, it may be important to prevent any further changes in gene expression after the tissue or cells has been removed from the subject. Changes in expression levels are known to change rapidly following perturbations, e.g., heat shock or activation with lipopolysaccharide (LPS) or other reagents. In addition, the RNA in the tissue and cells may quickly become degraded. Accordingly, in a preferred embodiment, the tissue or cells obtained from a subject is snap frozen as soon as possible.

RNA can be extracted from the tissue sample by a variety of methods, e.g., the guanidium thiocyanate lysis followed by CsCl centrifugation (Chirgwin et al. (1979) Biochemistry 18:5294-5299). RNA from single cells can be obtained as described in methods for preparing cDNA libraries from single cells, such as those described in Dulac, C. (1998) Curr. Top. Dev. Biol. 36:245 and Jena et al. (1996) J. Immunol. Methods 190: 199. Care to avoid RNA degradation must be taken, e.g., by inclusion of RNAsin. The RNA sample can then be enriched in particular species. In one embodiment, poly(A)+ RNA is isolated from the RNA sample. In general, such purification takes advantage of the poly-A tails on mRNA. In particular and as noted above, poly-T oligonucleotides may be immobilized within on a solid support to serve as affinity ligands for mRNA. Kits for this purpose are commercially available, e.g., the MessageMaker kit (Life Technologies, Grand Island, NY).

In a preferred embodiment, the RNA population is enriched in marker sequences. Enrichment can be undertaken, e.g., by primer-specific cDNA synthesis, or multiple rounds of linear amplification based on cDNA synthesis and template-directed in vitro

transcription {see, e.g., Wang et al. (1989) PNAS 86, 9717; Dulac et al., supra, and Jena et al., supra).

The population of RNA, enriched or not in particular species or sequences, can further be amplified. As defined herein, an "amplification process" is designed to strengthen, increase, or augment a molecule within the RNA. For example, where RNA is mRNA, an amplification process such as RT-PCR can be utilized to amplify the mRNA, such that a signal is detectable or detection is enhanced. Such an amplification process is beneficial particularly when the biological, tissue, or tumor sample is of a small size or volume.

Various amplification and detection methods can be used. For example, it is within the scope of the present invention to reverse transcribe mRNA into cDNA followed by polymerase chain reaction (RT-PCR); or, to use a single enzyme for both steps as described in U.S. Pat. No. 5,322,770, or reverse transcribe mRNA into cDNA followed by symmetric gap ligase chain reaction (RT-AGLCR) as described by R. L. Marshall, et al., PCR

Methods and Applications 4: 80-84 (1994). Real time PCR may also be used.

Other known amplification methods which can be utilized herein include but are not limited to the so-called "NASBA" or "3SR" technique described m PNAS USA 87: 1874- 1878 (1990) and also described in Nature 350 (No. 6313): 91-92 (1991); Q-beta

amplification as described in published European Patent Application (EPA) No. 4544610; strand displacement amplification (as described in G. T. Walker et al., Clin. Chem. 42: 9-13 (1996) and European Patent Application No. 684315; target mediated amplification, as described by PCT Publication W09322461; PCR; ligase chain reaction (LCR) (see, e.g., Wu and Wallace, Genomics 4, 560 (1989), Landegren et al., Science 241, 1077 (1988)); self-sustained sequence replication (SSR) (see, e.g., Guatelli et al., Proc. Nat. Acad. Sci. USA, 87, 1874 (1990)); and transcription amplification (see, e.g., Kwoh et al., Proc. Natl. Acad. Sci. USA 86, 1173 (1989)).

Many techniques are known in the state of the art for determining absolute and relative levels of gene expression, commonly used techniques suitable for use in the present invention include Northern analysis, RNase protection assays (RPA), microarrays and PCR- based techniques, such as quantitative PCR and differential display PCR. For example, Northern blotting involves running a preparation of RNA on a denaturing agarose gel, and transferring it to a suitable support, such as activated cellulose, nitrocellulose or glass or nylon membranes. Radiolabeled cDNA or RNA is then hybridized to the preparation, washed and analyzed by autoradiography.

In situ hybridization visualization may also be employed, wherein a radioactively labeled antisense RNA probe is hybridized with a thin section of a biopsy sample, washed, cleaved with RNase and exposed to a sensitive emulsion for autoradiography. The samples may be stained with hematoxylin to demonstrate the histological composition of the sample, and dark field imaging with a suitable light filter shows the developed emulsion. Non-radioactive labels such as digoxigenin may also be used.

Alternatively, mRNA expression can be detected on a DNA array, chip or a microarray. Labeled nucleic acids of a test sample obtained from a subject may be hybridized to a solid surface comprising biomarker DNA. Positive hybridization signal is obtained with the sample containing biomarker transcripts. Methods of preparing DNA arrays and their use are well known in the art (see, e.g., U.S. Pat. Nos: 6,618,6796;

6,379,897; 6,664,377; 6,451,536; 548,257; U.S. 20030157485 and Schena et al. (1995) Science 20, 467-470; Gerhold et al. (1999) Trends In Biochem. Sci. 24, 168-173; and Lennon et al. (2000) Drug Discovery Today 5, 59-65, which are herein incorporated by reference in their entirety). Serial Analysis of Gene Expression (SAGE) can also be performed (See for example U.S. Patent Application 20030215858).

To monitor mRNA levels, for example, mRNA is extracted from the biological sample to be tested, reverse transcribed, and fluorescently-labeled cDNA probes are generated. The microarrays capable of hybridizing to marker cDNA are then probed with the labeled cDNA probes, the slides scanned and fluorescence intensity measured. This intensity correlates with the hybridization intensity and expression levels.

Types of probes that can be used in the methods described herein include cDNA, riboprobes, synthetic oligonucleotides and genomic probes. The type of probe used will generally be dictated by the particular situation, such as riboprobes for in situ hybridization, and cDNA for Northern blotting, for example. In one embodiment, the probe is directed to nucleotide regions unique to the RNA. The probes may be as short as is required to differentially recognize marker mRNA transcripts, and may be as short as, for example, 15 bases; however, probes of at least 17, 18, 19 or 20 or more bases can be used. In one embodiment, the primers and probes hybridize specifically under stringent conditions to a DNA fragment having the nucleotide sequence corresponding to the marker. As herein used, the term "stringent conditions" means hybridization will occur only if there is at least 95% identity in nucleotide sequences. In another embodiment, hybridization under "stringent conditions" occurs when there is at least 97% identity between the sequences.

The form of labeling of the probes may be any that is appropriate, such as the use of radioisotopes, for example, ³² P and ³⁵ S. Labeling with radioisotopes may be achieved, whether the probe is synthesized chemically or biologically, by the use of suitably labeled bases.

In one embodiment, the biological sample contains polypeptide molecules from the test subject. Alternatively, the biological sample can contain mRNA molecules from the test subject or genomic DNA molecules from the test subject.

In another embodiment, the methods further involve obtaining a control biological sample from a control subject, contacting the control sample with a compound or agent capable of detecting marker polypeptide, mRNA, genomic DNA, or fragments thereof, such that the presence of the marker polypeptide, mRNA, genomic DNA, or fragments thereof, is detected in the biological sample, and comparing the presence of the marker polypeptide, mRNA, genomic DNA, or fragments thereof, in the control sample with the presence of the marker polypeptide, mRNA, genomic DNA, or fragments thereof in the test sample.

c. Methods for Detection of Biomarker Protein Expression

The activity or level of a biomarker protein can be detected and/or quantified by detecting or quantifying the expressed polypeptide. The polypeptide can be detected and quantified by any of a number of means well known to those of skill in the art. Aberrant levels of polypeptide expression of the polypeptides encoded by a biomarker nucleic acid and functionally similar homologs thereof, including a fragment or genetic alteration thereof {e.g., in regulatory or promoter regions thereof) are associated with the likelihood of response of a cancer to an immune checkpoint therapy. Any method known in the art for detecting polypeptides can be used. Such methods include, but are not limited to, immunodiffusion, Immunoelectrophoresis, radioimmunoassay (RIA), enzyme-linked immunosorbent assays (ELISAs), immunofluorescent assays, Western blotting, binder- ligand assays, immunohistochemical techniques, agglutination, complement assays, high performance liquid chromatography (HPLC), thin layer chromatography (TLC), hyperdiffusion chromatography, and the like (e.g., Basic and Clinical Immunology, Sites and Terr, eds., Appleton and Lange, Norwalk, Conn, pp 217-262, 1991 which is incorporated by reference). Preferred are binder-ligand immunoassay methods including reacting antibodies with an epitope or epitopes and competitively displacing a labeled polypeptide or derivative thereof.

For example, ELISA and RIA procedures may be conducted such that a desired biomarker protein standard is labeled (with a radioisotope such as ¹²⁵ I or ³⁵ S, or an assayable enzyme, such as horseradish peroxidase or alkaline phosphatase), and, together with the unlabelled sample, brought into contact with the corresponding antibody, whereon a second antibody is used to bind the first, and radioactivity or the immobilized enzyme assayed (competitive assay). Alternatively, the biomarker protein in the sample is allowed to react with the corresponding immobilized antibody, radioisotope- or enzyme-labeled anti-biomarker proteinantibody is allowed to react with the system, and radioactivity or the enzyme assayed (ELISA-sandwich assay). Other conventional methods may also be employed as suitable.

The above techniques may be conducted essentially as a "one-step" or "two-step" assay. A "one-step" assay involves contacting antigen with immobilized antibody and, without washing, contacting the mixture with labeled antibody. A "two-step" assay involves washing before contacting, the mixture with labeled antibody. Other conventional methods may also be employed as suitable.

In one embodiment, a method for measuring biomarker protein levels comprises the steps of: contacting a biological specimen with an antibody or variant (e.g., fragment) thereof which selectively binds the biomarker protein, and detecting whether said antibody or variant thereof is bound to said sample and thereby measuring the levels of the biomarker protein.

Enzymatic and radiolabeling of biomarker protein and/or the antibodies may be effected by conventional means. Such means will generally include covalent linking of the enzyme to the antigen or the antibody in question, such as by glutaraldehyde, specifically so as not to adversely affect the activity of the enzyme, by which is meant that the enzyme must still be capable of interacting with its substrate, although it is not necessary for all of the enzyme to be active, provided that enough remains active to permit the assay to be effected. Indeed, some techniques for binding enzyme are non-specific (such as using formaldehyde), and will only yield a proportion of active enzyme.

It is usually desirable to immobilize one component of the assay system on a support, thereby allowing other components of the system to be brought into contact with the component and readily removed without laborious and time-consuming labor. It is possible for a second phase to be immobilized away from the first, but one phase is usually sufficient.

It is possible to immobilize the enzyme itself on a support, but if solid-phase enzyme is required, then this is generally best achieved by binding to antibody and affixing the antibody to a support, models and systems for which are well-known in the art. Simple polyethylene may provide a suitable support.

Enzymes employable for labeling are not particularly limited, but may be selected from the members of the oxidase group, for example. These catalyze production of hydrogen peroxide by reaction with their substrates, and glucose oxidase is often used for its good stability, ease of availability and cheapness, as well as the ready availability of its substrate (glucose). Activity of the oxidase may be assayed by measuring the concentration of hydrogen peroxide formed after reaction of the enzyme-labeled antibody with the substrate under controlled conditions well-known in the art.

Other techniques may be used to detect biomarker protein according to a practitioner's preference based upon the present disclosure. One such technique is Western blotting (Towbin et at., Proc. Nat. Acad. Sci. 76:4350 (1979)), wherein a suitably treated sample is run on an SDS-PAGE gel before being transferred to a solid support, such as a nitrocellulose filter. Anti-biomarker protein antibodies (unlabeled) are then brought into contact with the support and assayed by a secondary immunological reagent, such as labeled protein A or anti-immunoglobulin (suitable labels including ¹²⁵ I, horseradish peroxidase and alkaline phosphatase). Chromatographic detection may also be used.

Immunohistochemistry may be used to detect expression of biomarker protein, e.g., in a biopsy sample. A suitable antibody is brought into contact with, for example, a thin layer of cells, washed, and then contacted with a second, labeled antibody. Labeling may be by fluorescent markers, enzymes, such as peroxidase, avidin, or radiolabelling. The assay is scored visually, using microscopy. Anti-biomarker protein antibodies, such as intrabodies, may also be used for imaging purposes, for example, to detect the presence of biomarker protein in cells and tissues of a subject. Suitable labels include radioisotopes, iodine ( ¹²⁵ I, ¹²¹ I), carbon ( ¹⁴ C), sulphur ( ³⁵ S), tritium ( ³ H), indium ( ¹¹² In), and technetium ( ⁹⁹ mTc), fluorescent labels, such as fluorescein and rhodamine, and biotin.

For in vivo imaging purposes, antibodies are not detectable, as such, from outside the body, and so must be labeled, or otherwise modified, to permit detection. Markers for this purpose may be any that do not substantially interfere with the antibody binding, but which allow external detection. Suitable markers may include those that may be detected by X-radiography, NMR or MRI. For X-radiographic techniques, suitable markers include any radioisotope that emits detectable radiation but that is not overtly harmful to the subject, such as barium or cesium, for example. Suitable markers for NMR and MRI generally include those with a detectable characteristic spin, such as deuterium, which may be incorporated into the antibody by suitable labeling of nutrients for the relevant hybridoma, for example.

The size of the subject, and the imaging system used, will determine the quantity of imaging moiety needed to produce diagnostic images. In the case of a radioisotope moiety, for a human subject, the quantity of radioactivity injected will normally range from about 5 to 20 millicuries of technetium-99. The labeled antibody or antibody fragment will then preferentially accumulate at the location of cells which contain biomarker protein. The labeled antibody or antibody fragment can then be detected using known techniques.

Antibodies that may be used to detect biomarker protein include any antibody, whether natural or synthetic, full length or a fragment thereof, monoclonal or polyclonal, that binds sufficiently strongly and specifically to the biomarker protein to be detected. An antibody may have a K _d of at most about 10 ^"6 M, 10 ^"7 M, 10 ^"8 M, 10 ^"9 M, 10 ^"10 M, 10 ^"U M, 10 ^{" 12} M. The phrase "specifically binds" refers to binding of, for example, an antibody to an epitope or antigen or antigenic determinant in such a manner that binding can be displaced or competed with a second preparation of identical or similar epitope, antigen or antigenic determinant. An antibody may bind preferentially to the biomarker protein relative to other proteins, such as related proteins.

Antibodies are commercially available or may be prepared according to methods known in the art. Antibodies and derivatives thereof that may be used encompass polyclonal or monoclonal antibodies, chimeric, human, humanized, primatized (CDR-grafted), veneered or single-chain antibodies as well as functional fragments, i.e., biomarker protein binding fragments, of antibodies. For example, antibody fragments capable of binding to a biomarker protein or portions thereof, including, but not limited to, Fv, Fab, Fab' and F(ab') 2 fragments can be used. Such fragments can be produced by enzymatic cleavage or by recombinant techniques. For example, papain or pepsin cleavage can generate Fab or F(ab') 2 fragments, respectively. Other proteases with the requisite substrate specificity can also be used to generate Fab or F(ab') 2 fragments. Antibodies can also be produced in a variety of truncated forms using antibody genes in which one or more stop codons have been introduced upstream of the natural stop site. For example, a chimeric gene encoding a F(ab') 2 heavy chain portion can be designed to include DNA sequences encoding the CH, domain and hinge region of the heavy chain.

Synthetic and engineered antibodies are described in, e.g., Cabilly et al., U.S. Pat. No. 4,816,567 Cabilly et al., European Patent No. 0,125,023 B l; Boss et al., U.S. Pat. No. 4,816,397; Boss et al., European Patent No. 0, 120,694 Bl; Neuberger, M. S. et al., WO 86/01533; Neuberger, M. S. et al., European Patent No. 0, 194,276 Bl; Winter, U.S. Pat. No. 5,225,539; Winter, European Patent No. 0,239,400 B l; Queen et al., European Patent No. 0451216 Bl; and Padlan, E. A. et al., EP 0519596 Al . See also, Newman, R. et al., BioTechnology, 10: 1455-1460 (1992), regarding primatized antibody, and Ladner et al., U.S. Pat. No. 4,946,778 and Bird, R. E. et al., Science, 242: 423-426 (1988)) regarding single-chain antibodies. Antibodies produced from a library, e.g., phage display library, may also be used.

In some embodiments, agents that specifically bind to a biomarker protein other than antibodies are used, such as peptides. Peptides that specifically bind to a biomarker protein can be identified by any means known in the art. For example, specific peptide binders of a biomarker protein can be screened for using peptide phage display libraries, d. Methods for Detection of Biomarker Structural Alterations

The following illustrative methods can be used to identify the presence of a structural alteration in a biomarker nucleic acid and/or biomarker polypeptide molecule in order to, for example, identify PBRM1 proteins that having mutations such as described herein. In certain embodiments, detection of the alteration involves the use of a

probe/primer in a polymerase chain reaction (PCR) (see, e.g., U.S. Pat. Nos. 4,683, 195 and 4,683,202), such as anchor PCR or RACE PCR, or, alternatively, in a ligation chain reaction (LCR) (see, e.g., Landegran et al. (1988) Science 241 : 1077-1080; and Nakazawa et al. (1994) Proc. Natl. Acad. Sci. USA 91 :360-364), the latter of which can be particularly useful for detecting point mutations in a biomarker nucleic acid such as a biomarker gene (see Abravaya et al. (1995) Nucleic Acids Res. 23 :675-682). This method can include the steps of collecting a sample of cells from a subject, isolating nucleic acid {e.g., genomic, mRNA or both) from the cells of the sample, contacting the nucleic acid sample with one or more primers which specifically hybridize to a biomarker gene under conditions such that hybridization and amplification of the biomarker gene (if present) occurs, and detecting the presence or absence of an amplification product, or detecting the size of the amplification product and comparing the length to a control sample. It is anticipated that PCR and/or LCR may be desirable to use as a preliminary amplification step in conjunction with any of the techniques used for detecting mutations described herein.

Alternative amplification methods include: self sustained sequence replication (Guatelli, J. C. et al. (1990) Proc. Natl. Acad. Sci. USA 87: 1874-1878), transcriptional amplification system (Kwoh, D. Y. et al. (1989) Proc. Natl. Acad. Sci. USA 86: 1173-1177), Q-Beta Replicase (Lizardi, P. M. et al. (1988) Bio-Technology 6: 1197), or any other nucleic acid amplification method, followed by the detection of the amplified molecules using techniques well known to those of skill in the art. These detection schemes are especially useful for the detection of nucleic acid molecules if such molecules are present in very low numbers.

In an alternative embodiment, mutations in a biomarker nucleic acid from a sample cell can be identified by alterations in restriction enzyme cleavage patterns. For example, sample and control DNA is isolated, amplified (optionally), digested with one or more restriction endonucleases, and fragment length sizes are determined by gel electrophoresis and compared. Differences in fragment length sizes between sample and control DNA indicates mutations in the sample DNA. Moreover, the use of sequence specific ribozymes (see, for example, U.S. Pat. No. 5,498,531) can be used to score for the presence of specific mutations by development or loss of a ribozyme cleavage site.

In other embodiments, genetic mutations in biomarker nucleic acid can be identified by hybridizing a sample and control nucleic acids, e.g., DNA or RNA, to high density arrays containing hundreds or thousands of oligonucleotide probes (Cronin, M. T. et al. (1996) Hum. Mutat. 7:244-255; Kozal, M. J. et al. (1996) Nat. Med. 2:753-759). For example, biomarker genetic mutations can be identified in two dimensional arrays containing light-generated DNA probes as described in Cronin et al. (1996) supra. Briefly, a first hybridization array of probes can be used to scan through long stretches of DNA in a sample and control to identify base changes between the sequences by making linear arrays of sequential, overlapping probes. This step allows the identification of point mutations. This step is followed by a second hybridization array that allows the characterization of specific mutations by using smaller, specialized probe arrays complementary to all variants or mutations detected. Each mutation array is composed of parallel probe sets, one complementary to the wild-type gene and the other complementary to the mutant gene. Such biomarker genetic mutations can be identified in a variety of contexts, including, for example, germline and somatic mutations.

In yet another embodiment, any of a variety of sequencing reactions known in the art can be used to directly sequence a biomarker gene and detect mutations by comparing the sequence of the sample biomarker with the corresponding wild-type (control) sequence. Examples of sequencing reactions include those based on techniques developed by Maxam and Gilbert (1977) Proc. Natl. Acad. Sci. USA 74:560 or Sanger (1977) Proc. Natl. Acad Sci. USA 74:5463. It is also contemplated that any of a variety of automated sequencing procedures can be utilized when performing the diagnostic assays (Naeve (1995)

Biotechniques 19:448-53), including sequencing by mass spectrometry (see, e.g., PCT International Publication No. WO 94/16101; Cohen et al. (1996) Adv. Chromatogr. 36: 127- 162; and Griffin et al. (1993) Appl. Biochem. Biotechnol. 38: 147-159).

Other methods for detecting mutations in a biomarker gene include methods in which protection from cleavage agents is used to detect mismatched bases in RNA/RNA or RNA/DNA heteroduplexes (Myers et al. (1985) Science 230: 1242). In general, the art technique of "mismatch cleavage" starts by providing heteroduplexes formed by hybridizing (labeled) RNA or DNA containing the wild-type biomarker sequence with potentially mutant RNA or DNA obtained from a tissue sample. The double-stranded duplexes are treated with an agent which cleaves single-stranded regions of the duplex such as which will exist due to base pair mismatches between the control and sample strands. For instance, RNA/DNA duplexes can be treated with RNase and DNA/DNA hybrids treated with SI nuclease to enzymatically digest the mismatched regions. In other embodiments, either DNA/DNA or RNA/DNA duplexes can be treated with hydroxylamine or osmium tetroxide and with piperidine in order to digest mismatched regions. After digestion of the mismatched regions, the resulting material is then separated by size on denaturing polyacrylamide gels to determine the site of mutation. See, for example, Cotton et al. (1988) Proc. Natl. Acad. Sci. USA 85:4397 and Saleeba et al. (1992) Methods Enzymol. 217:286-295. In a preferred embodiment, the control DNA or RNA can be labeled for detection.

In still another embodiment, the mismatch cleavage reaction employs one or more proteins that recognize mismatched base pairs in double-stranded DNA (so called "DNA mismatch repair" enzymes) in defined systems for detecting and mapping point mutations in biomarker cDNAs obtained from samples of cells. For example, the mutY enzyme of E. coli cleaves A at G/A mismatches and the thymidine DNA glycosylase from HeLa cells cleaves T at G/T mismatches (Hsu et al. (1994) Carcinogenesis 15: 1657-1662). According to an exemplary embodiment, a probe based on a biomarker sequence, e.g., a wild-type biomarker treated with a DNA mismatch repair enzyme, and the cleavage products, if any, can be detected from electrophoresis protocols or the like {e.g., U.S. Pat. No. 5,459,039.)

In other embodiments, alterations in electrophoretic mobility can be used to identify mutations in biomarker genes. For example, single strand conformation polymorphism (SSCP) may be used to detect differences in electrophoretic mobility between mutant and wild type nucleic acids (Orita et al. (1989) Proc Natl. Acad. Sci USA 86:2766; see also

Cotton (1993) Mutat. Res. 285: 125-144 and Hayashi (1992) Genet. Anal. Tech. Appl. 9:73- 79). Single-stranded DNA fragments of sample and control biomarker nucleic acids will be denatured and allowed to renature. The secondary structure of single-stranded nucleic acids varies according to sequence, the resulting alteration in electrophoretic mobility enables the detection of even a single base change. The DNA fragments may be labeled or detected with labeled probes. The sensitivity of the assay may be enhanced by using RNA (rather than DNA), in which the secondary structure is more sensitive to a change in sequence. In a preferred embodiment, the subject method utilizes heteroduplex analysis to separate double stranded heteroduplex molecules on the basis of changes in electrophoretic mobility (Keen et al. ( 1991 ) Trends Genet. 7:5).

In yet another embodiment the movement of mutant or wild-type fragments in polyacrylamide gels containing a gradient of denaturant is assayed using denaturing gradient gel electrophoresis (DGGE) (Myers et al. (1985) Nature 313 :495). When DGGE is used as the method of analysis, DNA will be modified to ensure that it does not completely denature, for example by adding a GC clamp of approximately 40 bp of high- melting GC-rich DNA by PCR. In a further embodiment, a temperature gradient is used in place of a denaturing gradient to identify differences in the mobility of control and sample DNA (Rosenbaum and Reissner (1987) Biophys. Chem. 265: 12753).

Examples of other techniques for detecting point mutations include, but are not limited to, selective oligonucleotide hybridization, selective amplification, or selective primer extension. For example, oligonucleotide primers may be prepared in which the known mutation is placed centrally and then hybridized to target DNA under conditions which permit hybridization only if a perfect match is found (Saiki et al. (1986) Nature 324: 163; Saiki et al. (1989) Proc. Natl. Acad. Sci. USA 86:6230). Such allele specific oligonucleotides are hybridized to PCR amplified target DNA or a number of different mutations when the oligonucleotides are attached to the hybridizing membrane and hybridized with labeled target DNA.

Alternatively, allele specific amplification technology which depends on selective

PCR amplification may be used in conjunction with the instant invention. Oligonucleotides used as primers for specific amplification may carry the mutation of interest in the center of the molecule (so that amplification depends on differential hybridization) (Gibbs et al. (1989) Nucleic Acids Res. 17:2437-2448) or at the extreme 3' end of one primer where, under appropriate conditions, mismatch can prevent, or reduce polymerase extension (Prossner (1993) Tibtech 11 :238). In addition it may be desirable to introduce a novel restriction site in the region of the mutation to create cleavage-based detection (Gasparini et al. (1992) Mol. Cell Probes 6: 1). It is anticipated that in certain embodiments amplification may also be performed using Taq ligase for amplification (Barany (1991) Proc. Natl. Acad. Sci USA 88: 189). In such cases, ligation will occur only if there is a perfect match at the 3' end of the 5' sequence making it possible to detect the presence of a known mutation at a specific site by looking for the presence or absence of amplification.

3. Anti-Cancer Therapies

The efficacy of immune checkpoint therapy is predicted according to biomarker amount and/or activity associated with a cancer in a subject according to the methods described herein. In one embodiment, such immune checkpoint therapy or combinations of therapies (e.g., anti-PD-1 antibodies) can be administered once a subject is indicated as being a likely responder to immune checkpoint therapy. In another embodiment, such immune checkpoint therapy can be avoided once a subject is indicated as not being a likely responder to immune checkpoint therapy and an alternative treatment regimen, such as targeted and/or untargeted anti-cancer therapies can be administered. Combination therapies are also contemplated and can comprise, for example, one or more

chemotherapeutic agents and radiation, one or more chemotherapeutic agents and immunotherapy, or one or more chemotherapeutic agents, radiation and chemotherapy, each combination of which can be with immune checkpoint therapy.

The term "targeted therapy" refers to administration of agents that selectively interact with a chosen biomolecule to thereby treat cancer. For example, anti-PBRMl agents, such as therapeutic monoclonal blocking antibodies, which are well-known in the art and described above, can be used to target tumor microenvironments and cells expressing unwanted PBRM1. Similarly, nivolumab (Opdivo®) is a human IgG4 anti-PD- 1 monoclonal antibody that blocks PD-1 activity (see, for example, Wang et al. (2014) Cancer Immunol. Res. 2:846-856; Johnson et al. (2015) Ther. Adv. Med. Oncol. 7:97-106; and Sundar et al. (2015) Ther. Adv. Med. Oncol. 7:85-96).

Immunotherapy is one form of targeted therapy that may comprise, for example, the use of cancer vaccines and/or sensitized antigen presenting cells. For example, an oncolytic virus is a virus that is able to infect and lyse cancer cells, while leaving normal cells unharmed, making them potentially useful in cancer therapy. Replication of oncolytic viruses both facilitates tumor cell destruction and also produces dose amplification at the tumor site. They may also act as vectors for anticancer genes, allowing them to be specifically delivered to the tumor site. The immunotherapy can involve passive immunity for short-term protection of a host, achieved by the administration of pre-formed antibody directed against a cancer antigen or disease antigen {e.g., administration of a monoclonal antibody, optionally linked to a chemotherapeutic agent or toxin, to a tumor antigen).

Immunotherapy can also focus on using the cytotoxic lymphocyte-recognized epitopes of cancer cell lines. Alternatively, antisense polynucleotides, ribozymes, RNA interference molecules, triple helix polynucleotides and the like, can be used to selectively modulate biomolecules that are linked to the initiation, progression, and/or pathology of a tumor or cancer.

The term "untargeted therapy" referes to administration of agents that do not selectively interact with a chosen biomolecule yet treat cancer. Representative examples of untargeted therapies include, without limitation, chemotherapy, gene therapy, and radiation therapy.

In one embodiment, chemotherapy is used. Chemotherapy includes the

administration of a chemotherapeutic agent. Such a chemotherapeutic agent may be, but is not limited to, those selected from among the following groups of compounds: platinum compounds, cytotoxic antibiotics, antimetabolities, anti-mitotic agents, alkylating agents, arsenic compounds, DNA topoisomerase inhibitors, taxanes, nucleoside analogues, plant alkaloids, and toxins; and synthetic derivatives thereof. Exemplary compounds include, but are not limited to, alkylating agents: cisplatin, treosulfan, and trofosfamide; plant alkaloids: vinblastine, paclitaxel, docetaxol; DNA topoisomerase inhibitors: teniposide, crisnatol, and mitomycin; anti -folates: methotrexate, mycophenolic acid, and hydroxyurea; pyrimidine analogs: 5-fluorouracil, doxifluridine, and cytosine arabinoside; purine analogs:

mercaptopurine and thioguanine; DNA antimetabolites: 2'-deoxy-5-fluorouridine, aphidicolin glycinate, and pyrazoloimidazole; and antimitotic agents: halichondrin, colchicine, and rhizoxin. Compositions comprising one or more chemotherapeutic agents (e.g., FLAG, CHOP) may also be used. FLAG comprises fludarabine, cytosine arabinoside (Ara-C) and G-CSF. CHOP comprises cyclophosphamide, vincristine, doxorubicin, and prednisone. In another embodiments, PARP (e.g., PARP-1 and/or PARP-2) inhibitors are used and such inhibitors are well known in the art (e.g., Olaparib, ABT-888, BSI-201, BGP-15 (N-Gene Research Laboratories, Inc.); INO-1001 (Inotek Pharmaceuticals Inc.); PJ34 (Soriano et al., 2001; Pacher et al, 2002b); 3-aminobenzamide (Trevigen); 4-amino- 1,8-naphthalimide; (Trevigen); 6(5H)-phenanthridinone (Trevigen); benzamide (U.S. Pat. Re. 36,397); and NU1025 (Bowman et al.). The mechanism of action is generally related to the ability of PARP inhibitors to bind PARP and decrease its activity. PARP catalyzes the conversion of .beta. -nicotinamide adenine dinucleotide (NAD+) into nicotinamide and poly-ADP-ribose (PAR). Both poly (ADP-ribose) and PARP have been linked to regulation of transcription, cell proliferation, genomic stability, and carcinogenesis

(Bouchard V. J. et.al. Experimental Hematology, Volume 31, Number 6, June 2003, pp. 446-454(9); Herceg Z.; Wang Z.-Q. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis, Volume 477, Number 1, 2 Jun. 2001, pp. 97-110(14)).

Poly(ADP-ribose) polymerase 1 (PARPl) is a key molecule in the repair of DNA single- strand breaks (SSBs) (de Murcia J. et al. 1997. Proc Natl Acad Sci USA 94:7303-7307; Schreiber V, Dantzer F, Ame J C, de Murcia G (2006) Nat Rev Mol Cell Biol 7:517-528; Wang Z Q, et al. (1997) Genes Dev 1 1 :2347-2358). Knockout of SSB repair by inhibition of PARP1 function induces DNA double-strand breaks (DSBs) that can trigger synthetic lethality in cancer cells with defective homology-directed DSB repair (Bryant H E, et al. (2005) Nature 434:913-917; Farmer H, et al. (2005) Nature 434:917-921). The foregoing examples of chemotherapeutic agents are illustrative, and are not intended to be limiting.

In another embodiment, radiation therapy is used. The radiation used in radiation therapy can be ionizing radiation. Radiation therapy can also be gamma rays, X-rays, or proton beams. Examples of radiation therapy include, but are not limited to, external-beam radiation therapy, interstitial implantation of radioisotopes (1-125, palladium, iridium), radioisotopes such as strontium-89, thoracic radiation therapy, intraperitoneal P-32 radiation therapy, and/or total abdominal and pelvic radiation therapy. For a general overview of radiation therapy, see Hellman, Chapter 16: Principles of Cancer Management: Radiation Therapy, 6th edition, 2001, DeVita et al., eds., J. B. Lippencott Company, Philadelphia. The radiation therapy can be administered as external beam radiation or teletherapy wherein the radiation is directed from a remote source. The radiation treatment can also be administered as internal therapy or brachytherapy wherein a radioactive source is placed inside the body close to cancer cells or a tumor mass. Also encompassed is the use of photodynamic therapy comprising the administration of photosensitizers, such as hematoporphyrin and its derivatives, Vertoporfin (BPD-MA), phthalocyanine,

photosensitizer Pc4, demethoxy-hypocrellin A; and 2B A-2-DMHA.

In another embodiment, hormone therapy is used. Hormonal therapeutic treatments can comprise, for example, hormonal agonists, hormonal antagonists (e.g., flutamide, bicalutamide, tamoxifen, raloxifene, leuprolide acetate (LUPRON), LH-RH antagonists), inhibitors of hormone biosynthesis and processing, and steroids (e.g., dexamethasone, retinoids, deltoids, betamethasone, Cortisol, cortisone, prednisone, dehydrotestosterone, glucocorticoids, mineralocorticoids, estrogen, testosterone, progestins), vitamin A derivatives (e.g., all-trans retinoic acid (ATRA)); vitamin D3 analogs; antigestagens (e.g., mifepristone, onapristone), or antiandrogens (e.g., cyproterone acetate).

In another embodiment, hyperthermia, a procedure in which body tissue is exposed to high temperatures (up to 106°F.) is used. Heat may help shrink tumors by damaging cells or depriving them of substances they need to live. Hyperthermia therapy can be local, regional, and whole-body hyperthermia, using external and internal heating devices.

Hyperthermia is almost always used with other forms of therapy (e.g., radiation therapy, chemotherapy, and biological therapy) to try to increase their effectiveness. Local hyperthermia refers to heat that is applied to a very small area, such as a tumor. The area may be heated externally with high-frequency waves aimed at a tumor from a device outside the body. To achieve internal heating, one of several types of sterile probes may be used, including thin, heated wires or hollow tubes filled with warm water; implanted microwave antennae; and radiofrequency electrodes. In regional hyperthermia, an organ or a limb is heated. Magnets and devices that produce high energy are placed over the region to be heated. In another approach, called perfusion, some of the patient's blood is removed, heated, and then pumped (perfused) into the region that is to be heated internally. Whole- body heating is used to treat metastatic cancer that has spread throughout the body. It can be accomplished using warm-water blankets, hot wax, inductive coils (like those in electric blankets), or thermal chambers (similar to large incubators). Hyperthermia does not cause any marked increase in radiation side effects or complications. Heat applied directly to the skin, however, can cause discomfort or even significant local pain in about half the patients treated. It can also cause blisters, which generally heal rapidly.

In still another embodiment, photodynamic therapy (also called PDT, photoradiation therapy, phototherapy, or photochemotherapy) is used for the treatment of some types of cancer. It is based on the discovery that certain chemicals known as photosensitizing agents can kill one-celled organisms when the organisms are exposed to a particular type of light. PDT destroys cancer cells through the use of a fixed-frequency laser light in combination with a photosensitizing agent. In PDT, the photosensitizing agent is injected into the bloodstream and absorbed by cells all over the body. The agent remains in cancer cells for a longer time than it does in normal cells. When the treated cancer cells are exposed to laser light, the photosensitizing agent absorbs the light and produces an active form of oxygen that destroys the treated cancer cells. Light exposure must be timed carefully so that it occurs when most of the photosensitizing agent has left healthy cells but is still present in the cancer cells. The laser light used in PDT can be directed through a fiberoptic (a very thin glass strand). The fiber-optic is placed close to the cancer to deliver the proper amount of light. The fiber-optic can be directed through a bronchoscope into the lungs for the treatment of lung cancer or through an endoscope into the esophagus for the treatment of esophageal cancer. An advantage of PDT is that it causes minimal damage to healthy tissue. However, because the laser light currently in use cannot pass through more than about 3 centimeters of tissue (a little more than one and an eighth inch), PDT is mainly used to treat tumors on or just under the skin or on the lining of internal organs.

Photodynamic therapy makes the skin and eyes sensitive to light for 6 weeks or more after treatment. Patients are advised to avoid direct sunlight and bright indoor light for at least 6 weeks. If patients must go outdoors, they need to wear protective clothing, including sunglasses. Other temporary side effects of PDT are related to the treatment of specific areas and can include coughing, trouble swallowing, abdominal pain, and painful breathing or shortness of breath. In December 1995, the U.S. Food and Drug Administration (FDA) approved a photosensitizing agent called porfimer sodium, or Photofrin®, to relieve symptoms of esophageal cancer that is causing an obstruction and for esophageal cancer that cannot be satisfactorily treated with lasers alone. In January 1998, the FDA approved porfimer sodium for the treatment of early nonsmall cell lung cancer in patients for whom the usual treatments for lung cancer are not appropriate. The National Cancer Institute and other institutions are supporting clinical trials (research studies) to evaluate the use of photodynamic therapy for several types of cancer, including cancers of the bladder, brain, larynx, and oral cavity.

In yet another embodiment, laser therapy is used to harness high-intensity light to destroy cancer cells. This technique is often used to relieve symptoms of cancer such as bleeding or obstruction, especially when the cancer cannot be cured by other treatments. It may also be used to treat cancer by shrinking or destroying tumors. The term "laser" stands for light amplification by stimulated emission of radiation. Ordinary light, such as that from a light bulb, has many wavelengths and spreads in all directions. Laser light, on the other hand, has a specific wavelength and is focused in a narrow beam. This type of high- intensity light contains a lot of energy. Lasers are very powerful and may be used to cut through steel or to shape diamonds. Lasers also can be used for very precise surgical work, such as repairing a damaged retina in the eye or cutting through tissue (in place of a scalpel). Although there are several different kinds of lasers, only three kinds have gained wide use in medicine: Carbon dioxide (CO2) laser—This type of laser can remove thin layers from the skin's surface without penetrating the deeper layers. This technique is particularly useful in treating tumors that have not spread deep into the skin and certain precancerous conditions. As an alternative to traditional scalpel surgery, the CO2 laser is also able to cut the skin. The laser is used in this way to remove skin cancers.

Neodymium:yttrium-aluminum-garnet (Nd:YAG) laser— Light from this laser can penetrate deeper into tissue than light from the other types of lasers, and it can cause blood to clot quickly. It can be carried through optical fibers to less accessible parts of the body. This type of laser is sometimes used to treat throat cancers. Argon laser— This laser can pass through only superficial layers of tissue and is therefore useful in dermatology and in eye surgery. It also is used with light-sensitive dyes to treat tumors in a procedure known as photodynamic therapy (PDT). Lasers have several advantages over standard surgical tools, including: Lasers are more precise than scalpels. Tissue near an incision is protected, since there is little contact with surrounding skin or other tissue. The heat produced by lasers sterilizes the surgery site, thus reducing the risk of infection. Less operating time may be needed because the precision of the laser allows for a smaller incision. Healing time is often shortened; since laser heat seals blood vessels, there is less bleeding, swelling, or scarring. Laser surgery may be less complicated. For example, with fiber optics, laser light can be directed to parts of the body without making a large incision. More procedures may be done on an outpatient basis. Lasers can be used in two ways to treat cancer: by shrinking or destroying a tumor with heat, or by activating a chemical—known as a photosensitizing agent— that destroys cancer cells. In PDT, a photosensitizing agent is retained in cancer cells and can be stimulated by light to cause a reaction that kills cancer cells. CO2 and Nd:YAG lasers are used to shrink or destroy tumors. They may be used with endoscopes, tubes that allow physicians to see into certain areas of the body, such as the bladder. The light from some lasers can be transmitted through a flexible endoscope fitted with fiber optics. This allows physicians to see and work in parts of the body that could not otherwise be reached except by surgery and therefore allows very precise aiming of the laser beam. Lasers also may be used with low-power microscopes, giving the doctor a clear view of the site being treated. Used with other instruments, laser systems can produce a cutting area as small as 200 microns in diameter— less than the width of a very fine thread. Lasers are used to treat many types of cancer. Laser surgery is a standard treatment for certain stages of glottis (vocal cord), cervical, skin, lung, vaginal, vulvar, and penile cancers. In addition to its use to destroy the cancer, laser surgery is also used to help relieve symptoms caused by cancer (palliative care). For example, lasers may be used to shrink or destroy a tumor that is blocking a patient's trachea (windpipe), making it easier to breathe. It is also sometimes used for palliation in colorectal and anal cancer. Laser- induced interstitial thermotherapy (LITT) is one of the most recent developments in laser therapy. LITT uses the same idea as a cancer treatment called hyperthermia; that heat may help shrink tumors by damaging cells or depriving them of substances they need to live. In this treatment, lasers are directed to interstitial areas (areas between organs) in the body. The laser light then raises the temperature of the tumor, which damages or destroys cancer cells.

The duration and/or dose of treatment with anti-immune checkpoint therapies may vary according to the particular anti-immune checkpoint agent or combination thereof. An appropriate treatment time for a particular cancer therapeutic agent will be appreciated by the skilled artisan. The present invention contemplates the continued assessment of optimal treatment schedules for each cancer therapeutic agent, where the phenotype of the cancer of the subject as determined by the methods of the present invention is a factor in determining optimal treatment doses and schedules.

Any means for the introduction of a polynucleotide into mammals, human or non- human, or cells thereof may be adapted to the practice of this invention for the delivery of the various constructs of the present invention into the intended recipient. In one embodiment of the present invention, the DNA constructs are delivered to cells by transfection, i.e., by delivery of "naked" DNA or in a complex with a colloidal dispersion system. A colloidal system includes macromolecule complexes, nanocapsules,

microspheres, beads, and lipid-based systems including oil-in-water emulsions, micelles, mixed micelles, and liposomes. The preferred colloidal system of this invention is a lipid- complexed or liposome-formulated DNA. In the former approach, prior to formulation of DNA, e.g., with lipid, a plasmid containing a transgene bearing the desired DNA constructs may first be experimentally optimized for expression {e.g., inclusion of an intron in the 5' untranslated region and elimination of unnecessary sequences (Feigner, et al., Ann NY Acad Sci 126-139, 1995). Formulation of DNA, e.g. with various lipid or liposome materials, may then be effected using known methods and materials and delivered to the recipient mammal. See, e.g., Canonico et al, Am J Respir Cell Mol Biol 10:24-29, 1994; Tsan et al, Am J Physiol 268; Alton et al., Nat Genet. 5: 135-142, 1993 and U.S. patent No. 5,679,647 by Carson et al.

The targeting of liposomes can be classified based on anatomical and mechanistic factors. Anatomical classification is based on the level of selectivity, for example, organ- specific, cell-specific, and organelle-specific. Mechanistic targeting can be distinguished based upon whether it is passive or active. Passive targeting utilizes the natural tendency of liposomes to distribute to cells of the reticulo-endothelial system (RES) in organs, which contain sinusoidal capillaries. Active targeting, on the other hand, involves alteration of the liposome by coupling the liposome to a specific ligand such as a monoclonal antibody, sugar, glycolipid, or protein, or by changing the composition or size of the liposome in order to achieve targeting to organs and cell types other than the naturally occurring sites of localization.

The surface of the targeted delivery system may be modified in a variety of ways.

In the case of a liposomal targeted delivery system, lipid groups can be incorporated into the lipid bilayer of the liposome in order to maintain the targeting ligand in stable association with the liposomal bilayer. Various linking groups can be used for joining the lipid chains to the targeting ligand. Naked DNA or DNA associated with a delivery vehicle, e.g., liposomes, can be administered to several sites in a subject (see below).

Nucleic acids can be delivered in any desired vector. These include viral or non- viral vectors, including adenovirus vectors, adeno-associated virus vectors, retrovirus vectors, lentivirus vectors, and plasmid vectors. Exemplary types of viruses include HSV (herpes simplex virus), AAV (adeno associated virus), HIV (human immunodeficiency virus), BIV (bovine immunodeficiency virus), and MLV (murine leukemia virus). Nucleic acids can be administered in any desired format that provides sufficiently efficient delivery levels, including in virus particles, in liposomes, in nanoparticles, and complexed to polymers.

The nucleic acids encoding a protein or nucleic acid of interest may be in a plasmid or viral vector, or other vector as is known in the art. Such vectors are well known and any can be selected for a particular application. In one embodiment of the present invention, the gene delivery vehicle comprises a promoter and a demethylase coding sequence.

Preferred promoters are tissue-specific promoters and promoters which are activated by cellular proliferation, such as the thymidine kinase and thymidylate synthase promoters. Other preferred promoters include promoters which are activatable by infection with a virus, such as the a- and β-interferon promoters, and promoters which are activatable by a hormone, such as estrogen. Other promoters which can be used include the Moloney virus LTR, the CMV promoter, and the mouse albumin promoter. A promoter may be constitutive or inducible.

In another embodiment, naked polynucleotide molecules are used as gene delivery vehicles, as described in WO 90/11092 and U.S. Patent 5,580,859. Such gene delivery vehicles can be either growth factor DNA or RNA and, in certain embodiments, are linked to killed adenovirus. Curiel et al., Hum. Gene. Ther. 3 : 147-154, 1992. Other vehicles which can optionally be used include DNA-ligand (Wu et al., J. Biol. Chem.

264: 16985-16987, 1989), lipid-DNA combinations (Feigner et al., Proc. Natl. Acad. Sci. USA 84:7413 7417, 1989), liposomes (Wang et al., Proc. Natl. Acad. Sci. 84:7851-7855, 1987) and microprojectiles (Williams et al., Proc. Natl. Acad. Sci. 88:2726-2730, 1991).

A gene delivery vehicle can optionally comprise viral sequences such as a viral origin of replication or packaging signal. These viral sequences can be selected from viruses such as astrovirus, coronavirus, orthomyxovirus, papovavirus, paramyxovirus, parvovirus, picornavirus, poxvirus, retrovirus, togavirus or adenovirus. In a preferred embodiment, the growth factor gene delivery vehicle is a recombinant retroviral vector. Recombinant retroviruses and various uses thereof have been described in numerous references including, for example, Mann et al., Cell 33 : 153, 1983, Cane and Mulligan, Proc. Nat'l. Acad. Sci. USA 81 :6349, 1984, Miller et al., Human Gene Therapy 1 :5-14, 1990, U.S. Patent Nos. 4,405,712, 4,861,719, and 4,980,289, and PCT Application Nos. WO 89/02,468, WO 89/05,349, and WO 90/02,806. Numerous retroviral gene delivery vehicles can be utilized in the present invention, including for example those described in EP 0,415,731; WO 90/07936; WO 94/03622; WO 93/25698; WO 93/25234; U.S. Patent No. 5,219,740; WO 9311230; WO 9310218; Vile and Hart, Cancer Res. 53 :3860-3864, 1993; Vile and Hart, Cancer Res. 53 :962-967, 1993; Ram et al., Cancer Res. 53 :83-88, 1993; Takamiya et al., J. Neurosci. Res. 33 :493-503, 1992; Baba et al., J. Neurosurg.

79:729-735, 1993 (U.S. Patent No. 4,777,127, GB 2,200,651, EP 0,345,242 and

WO91/02805).

Other viral vector systems that can be used to deliver a polynucleotide of the present invention have been derived from herpes virus, e.g., Herpes Simplex Virus (U.S. Patent No. 5,631,236 by Woo et al., issued May 20, 1997 and WO 00/08191 by Neurovex), vaccinia virus (Ridgeway (1988) Ridgeway, "Mammalian expression vectors," In: Rodriguez R L, Denhardt D T, ed. Vectors: A survey of molecular cloning vectors and their uses.

Stoneham: Butterworth,; Baichwal and Sugden (1986) "Vectors for gene transfer derived from animal DNA viruses: Transient and stable expression of transferred genes," In:

Kucherlapati R, ed. Gene transfer. New York: Plenum Press; Coupar et al. (1988) Gene, 68: 1-10), and several RNA viruses. Preferred viruses include an alphavirus, a poxivirus, an arena virus, a vaccinia virus, a polio virus, and the like. They offer several attractive features for various mammalian cells (Friedmann (1989) Science, 244: 1275-1281; Ridgeway, 1988, supra; Baichwal and Sugden, 1986, supra; Coupar et al., 1988; Horwich et al.(1990) J.Virol., 64:642-650).

In other embodiments, target DNA in the genome can be manipulated using well- known methods in the art. For example, the target DNA in the genome can be manipulated by deletion, insertion, and/or mutation are retroviral insertion, artificial chromosome techniques, gene insertion, random insertion with tissue specific promoters, gene targeting, transposable elements and/or any other method for introducing foreign DNA or producing modified DNA/modified nuclear DNA. Other modification techniques include deleting DNA sequences from a genome and/or altering nuclear DNA sequences. Nuclear DNA sequences, for example, may be altered by site-directed mutagenesis.

In other embodiments, recombinant biomarker polypeptides, and fragments thereof, can be administered to subjects. In some embodiments, fusion proteins can be constructed and administered which have enhanced biological properties. In addition, the biomarker polypeptides, and fragment thereof, can be modified according to well-known

pharmacological methods in the art (e.g., pegylation, glycosylation, oligomerization, etc.) in order to further enhance desirable biological activities, such as increased bioavailability and decreased proteolytic degradation.

4. Clincal Efficacy

Clinical efficacy can be measured by any method known in the art. For example, the response to a therapy, such as anti-immune checkpoint therapies, relates to any response of the cancer, e.g., a tumor, to the therapy, preferably to a change in tumor mass and/or volume after initiation of neoadjuvant or adjuvant chemotherapy. Tumor response may be assessed in a neoadjuvant or adjuvant situation where the size of a tumor after systemic intervention can be compared to the initial size and dimensions as measured by CT, PET, mammogram, ultrasound or palpation and the cellularity of a tumor can be estimated histologically and compared to the cellularity of a tumor biopsy taken before initiation of treatment. Response may also be assessed by caliper measurement or pathological examination of the tumor after biopsy or surgical resection. Response may be recorded in a quantitative fashion like percentage change in tumor volume or cellularity or using a semiquantitative scoring system such as residual cancer burden (Symmans et al, J. Clin. Oncol. (2007) 25:4414-4422) or Miller-Payne score (Ogston et al, (2003) Breast (Edinburgh, Scotland) 12:320-327) in a qualitative fashion like "pathological complete response" (pCR), "clinical complete remission" (cCR), "clinical partial remission" (cPR), "clinical stable disease" (cSD), "clinical progressive disease" (cPD) or other qualitative criteria. Assessment of tumor response may be performed early after the onset of neoadjuvant or adjuvant therapy, e.g., after a few hours, days, weeks or preferably after a few months. A typical endpoint for response assessment is upon termination of neoadjuvant chemotherapy or upon surgical removal of residual tumor cells and/or the tumor bed.

In some embodiments, clinical efficacy of the therapeutic treatments described herein may be determined by measuring the clinical benefit rate (CBR). The clinical benefit rate is measured by determining the sum of the percentage of patients who are in complete remission (CR), the number of patients who are in partial remission (PR) and the number of patients having stable disease (SD) at a time point at least 6 months out from the end of therapy. The shorthand for this formula is CBR=CR+PR+SD over 6 months. In some embodiments, the CBR for a particular anti-immune checkpoint therapeutic regimen is at least 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, or more.

Additional criteria for evaluating the response to anti-immune checkpoint therapies are related to "survival," which includes all of the following: survival until mortality, also known as overall survival (wherein said mortality may be either irrespective of cause or tumor related); "recurrence-free survival" (wherein the term recurrence shall include both localized and distant recurrence); metastasis free survival; disease free survival (wherein the term disease shall include cancer and diseases associated therewith). The length of said survival may be calculated by reference to a defined start point (e.g., time of diagnosis or start of treatment) and end point (e.g., death, recurrence or metastasis). In addition, criteria for efficacy of treatment can be expanded to include response to chemotherapy, probability of survival, probability of metastasis within a given time period, and probability of tumor recurrence.

For example, in order to determine appropriate threshold values, a particular anti- immune checkpoint therapeutic regimen can be administered to a population of subjects and the outcome can be correlated to biomarker measurements that were determined prior to administration of any immune checkpoint therapy. The outcome measurement may be pathologic response to therapy given in the neoadjuvant setting. Alternatively, outcome measures, such as overall survival and disease-free survival can be monitored over a period of time for subjects following immune checkpoint therapy for whom biomarker measurement values are known. In certain embodiments, the same doses of anti-immune checkpoint agents are administered to each subject. In related embodiments, the doses administered are standard doses known in the art for anti-immune checkpoint agents. The period of time for which subjects are monitored can vary. For example, subjects may be monitored for at least 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 25, 30, 35, 40, 45, 50, 55, or 60 months. Biomarker measurement threshold values that correlate to outcome of an immune checkpoint therapy can be determined using methods such as those described in the Examples section. 5. Further Uses and Methods of the Present Invention

The methods described herein can be used in a variety of diagnostic, prognostic, and therapeutic applications. In any method described herein, such as a diagnostic method, prognostic method, therapeutic method, or combination thereof, all steps of the method can be performed by a single actor or, alternatively, by more than one actor. For example, diagnosis can be performed directly by the actor providing therapeutic treatment.

Alternatively, a person providing a therapeutic agent can request that a diagnostic assay be performed. The diagnostician and/or the therapeutic interventionist can interpret the diagnostic assay results to determine a therapeutic strategy. Similarly, such alternative processes can apply to other assays, such as prognostic assays. The compositions described herein can also be used in a variety of diagnostic, prognostic, and therapeutic applications regarding biomarkers described herein, such as those listed in Table 1. Moreover, any method of diagnosis, prognosis, prevention, and the like described herein can be be applied to a therapy or test agent of interest, such as immune checkpoint therapies, EGFR therapies, anti-cancer therapies, and the like. a. Screening Methods

One aspect of the present invention relates to screening assays, including non-cell based assays. In one embodiment, the assays provide a method for identifying whether a cancer is likely to respond to immune checkpoint therapy and/or whether an agent can inhibit the growth of or kill a cancer cell that is unlikely to respond to immune checkpoint therapy.

In one embodiment, the present invention relates to assays for screening test agents which bind to, or modulate the biological activity of, at least one biomarker listed in Table

- I l l - 1. In one embodiment, a method for identifying such an agent entails determining the ability of the agent to modulate, e.g. inhibit, the at least one biomarker listed in Table 1.

In one embodiment, an assay is a cell-free or cell-based assay, comprising contacting at least one biomarker listed in Table 1, with a test agent, and determining the ability of the test agent to modulate (e.g. inhibit) the enzymatic activity of the biomarker, such as by measuring direct binding of substrates or by measuring indirect parameters as described below.

For example, in a direct binding assay, biomarker protein (or their respective target polypeptides or molecules) can be coupled with a radioisotope or enzymatic label such that binding can be determined by detecting the labeled protein or molecule in a complex. For example, the targets can be labeled with ¹²⁵ 1, ³⁵ S, ¹⁴ C, or ³ H, either directly or indirectly, and the radioisotope detected by direct counting of radioemmission or by scintillation counting. Alternatively, the targets can be enzymatically labeled with, for example, horseradish peroxidase, alkaline phosphatase, or luciferase, and the enzymatic label detected by determination of conversion of an appropriate substrate to product.

Determining the interaction between biomarker and substrate can also be accomplished using standard binding or enzymatic analysis assays. In one or more embodiments of the above described assay methods, it may be desirable to immobilize polypeptides or molecules to facilitate separation of complexed from uncomplexed forms of one or both of the proteins or molecules, as well as to accommodate automation of the assay.

Binding of a test agent to a target can be accomplished in any vessel suitable for containing the reactants. Non-limiting examples of such vessels include microtiter plates, test tubes, and micro-centrifuge tubes. Immobilized forms of the antibodies of the present invention can also include antibodies bound to a solid phase like a porous, microporous (with an average pore diameter less than about one micron) or macroporous (with an average pore diameter of more than about 10 microns) material, such as a membrane, cellulose, nitrocellulose, or glass fibers; a bead, such as that made of agarose or polyacrylamide or latex; or a surface of a dish, plate, or well, such as one made of polystyrene.

In an alternative embodiment, determining the ability of the agent to modulate the interaction between the biomarker and a substrate or a biomarker and its natural binding partner can be accomplished by determining the ability of the test agent to modulate the activity of a polypeptide or other product that functions downstream or upstream of its position within the signaling pathway (e.g., feedback loops). Such feedback loops are well- known in the art (see, for example, Chen and Guillemin (2009) Int. J. Tryptophan Res. 2: 1- 19).

The present invention further pertains to novel agents identified by the above- described screening assays. Accordingly, it is within the scope of this invention to further use an agent identified as described herein in an appropriate animal model. For example, an agent identified as described herein can be used in an animal model to determine the efficacy, toxicity, or side effects of treatment with such an agent. Alternatively, an antibody identified as described herein can be used in an animal model to determine the mechanism of action of such an agent.

b. Predictive Medicine

The present invention also pertains to the field of predictive medicine in which diagnostic assays, prognostic assays, and monitoring clinical trials are used for prognostic (predictive) purposes to thereby treat an individual prophylactically. Accordingly, one aspect of the present invention relates to diagnostic assays for determining the amount and/or activity level of a biomarker listed in Table 1 in the context of a biological sample (e.g., blood, serum, cells, or tissue) to thereby determine whether an individual afflicted with a cancer is likely to respond to immune checkpoint therapy, whether in an original or recurrent cancer. Such assays can be used for prognostic or predictive purpose to thereby prophylactically treat an individual prior to the onset or after recurrence of a disorder characterized by or associated with biomarker polypeptide, nucleic acid expression or activity. The skilled artisan will appreciate that any method can use one or more (e.g., combinations) of biomarkers listed in Table 1.

Another aspect of the present invention pertains to monitoring the influence of agents (e.g., drugs, compounds, and small nucleic acid-based molecules) on the expression or activity of a biomarker listed in Table 1. These and other agents are described in further detail in the following sections.

The skilled artisan will also appreciated that, in certain embodiments, the methods of the present invention implement a computer program and computer system. For example, a computer program can be used to perform the algorithms described herein. A computer system can also store and manipulate data generated by the methods of the present invention which comprises a plurality of biomarker signal changes/profiles which can be used by a computer system in implementing the methods of this invention. In certain embodiments, a computer system receives biomarker expression data; (ii) stores the data; and (iii) compares the data in any number of ways described herein (e.g., analysis relative to appropriate controls) to determine the state of informative biomarkers from cancerous or pre-cancerous tissue. In other embodiments, a computer system (i) compares the determined expression biomarker level to a threshold value; and (ii) outputs an indication of whether said biomarker level is significantly modulated (e.g., above or below) the threshold value, or a phenotype based on said indication.

In certain embodiments, such computer systems are also considered part of the present invention. Numerous types of computer systems can be used to implement the analytic methods of this invention according to knowledge possessed by a skilled artisan in the bioinformatics and/or computer arts. Several software components can be loaded into memory during operation of such a computer system. The software components can comprise both software components that are standard in the art and components that are special to the present invention (e.g., dCHIP software described in Lin et al. (2004) Bioinformatics 20, 1233-1240; radial basis machine learning algorithms (RBM) known in the art).

The methods of the present invention can also be programmed or modeled in mathematical software packages that allow symbolic entry of equations and high-level specification of processing, including specific algorithms to be used, thereby freeing a user of the need to procedurally program individual equations and algorithms. Such packages include, e.g., Matlab from Mathworks (Natick, Mass.), Mathematica from Wolfram

Research (Champaign, 111.) or S-Plus from MathSoft (Seattle, Wash.).

In certain embodiments, the computer comprises a database for storage of biomarker data. Such stored profiles can be accessed and used to perform comparisons of interest at a later point in time. For example, biomarker expression profiles of a sample derived from the non-cancerous tissue of a subject and/or profiles generated from population-based distributions of informative loci of interest in relevant populations of the same species can be stored and later compared to that of a sample derived from the cancerous tissue of the subject or tissue suspected of being cancerous of the subject.

In addition to the exemplary program structures and computer systems described herein, other, alternative program structures and computer systems will be readily apparent to the skilled artisan. Such alternative systems, which do not depart from the above described computer system and programs structures either in spirit or in scope, are therefore intended to be comprehended within the accompanying claims,

c. Diagnostic Assays

The present invention provides, in part, methods, systems, and code for accurately classifying whether a biological sample is associated with a cancer that is likely to respond to immune checkpoint therapy. In some embodiments, the present invention is useful for classifying a sample (e.g., from a subject) as associated with or at risk for responding to or not responding to immune checkpoint therapy using a statistical algorithm and/or empirical data (e.g., the amount or activity of a biomarker listed in Table 1).

An exemplary method for detecting the amount or activity of a biomarker listed in

Table 1, and thus useful for classifying whether a sample is likely or unlikely to respond to immune checkpoint therapy involves obtaining a biological sample from a test subject and contacting the biological sample with an agent, such as a protein-binding agent like an antibody or antigen-binding fragment thereof, or a nucleic acid-binding agent like an oligonucleotide, capable of detecting the amount or activity of the biomarker in the biological sample. In some embodiments, at least one antibody or antigen-binding fragment thereof is used, wherein two, three, four, five, six, seven, eight, nine, ten, or more such antibodies or antibody fragments can be used in combination (e.g., in sandwich ELISAs) or in serial. In certain instances, the statistical algorithm is a single learning statistical classifier system. For example, a single learning statistical classifier system can be used to classify a sample as a based upon a prediction or probability value and the presence or level of the biomarker. The use of a single learning statistical classifier system typically classifies the sample as, for example, a likely immune checkpoint therapy responder or progressor sample with a sensitivity, specificity, positive predictive value, negative predictive value, and/or overall accuracy of at least about 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99%.

Other suitable statistical algorithms are well known to those of skill in the art. For example, learning statistical classifier systems include a machine learning algorithmic technique capable of adapting to complex data sets (e.g., panel of markers of interest) and making decisions based upon such data sets. In some embodiments, a single learning statistical classifier system such as a classification tree (e.g., random forest) is used. In other embodiments, a combination of 2, 3, 4, 5, 6, 7, 8, 9, 10, or more learning statistical classifier systems are used, preferably in tandem. Examples of learning statistical classifier systems include, but are not limited to, those using inductive learning (e.g.,

decision/classification trees such as random forests, classification and regression trees (C&RT), boosted trees, etc.), Probably Approximately Correct (PAC) learning,

connectionist learning (e.g., neural networks (NN), artificial neural networks (ANN), neuro fuzzy networks (NFN), network structures, perceptrons such as multi-layer perceptrons, multi-layer feed-forward networks, applications of neural networks, Bayesian learning in belief networks, etc.), reinforcement learning (e.g., passive learning in a known

environment such as naive learning, adaptive dynamic learning, and temporal difference learning, passive learning in an unknown environment, active learning in an unknown environment, learning action-value functions, applications of reinforcement learning, etc.), and genetic algorithms and evolutionary programming. Other learning statistical classifier systems include support vector machines (e.g., Kernel methods), multivariate adaptive regression splines (MARS), Levenberg-Marquardt algorithms, Gauss-Newton algorithms, mixtures of Gaussians, gradient descent algorithms, and learning vector quantization (LVQ). In certain embodiments, the method of the present invention further comprises sending the sample classification results to a clinician, e.g., an oncologist.

In another embodiment, the diagnosis of a subject is followed by administering to the individual a therapeutically effective amount of a defined treatment based upon the diagnosis.

In one embodiment, the methods further involve obtaining a control biological sample (e.g., biological sample from a subject who does not have a cancer or whose cancer is susceptible to immune checkpoint therapy), a biological sample from the subject during remission, or a biological sample from the subject during treatment for developing a cancer progressing despite immune checkpoint therapy.

d. Prognostic Assays

The diagnostic methods described herein can furthermore be utilized to identify subjects having or at risk of developing a cancer that is likely or unlikely to be responsive to immune checkpoint therapy. The assays described herein, such as the preceding diagnostic assays or the following assays, can be utilized to identify a subject having or at risk of developing a disorder associated with a misregulation of the amount or activity of at least one biomarker described in Table 1, such as in cancer. Alternatively, the prognostic assays can be utilized to identify a subject having or at risk for developing a disorder associated with a misregulation of the at least one biomarker described in Table 1, such as in cancer. Furthermore, the prognostic assays described herein can be used to determine whether a subject can be administered an agent (e.g., an agonist, antagonist,

peptidomimetic, polypeptide, peptide, nucleic acid, small molecule, or other drug candidate) to treat a disease or disorder associated with the aberrant biomarker expression or activity.

e. Treatment Methods

The compositions described herein (including dual binding antibodies and derivatives and conjugates thereof) can be used in a variety of in vitro and in vivo therapeutic applications using the formulations and/or combinations described herein. In one embodiment, anti-immune checkpoint agents can be used to treat cancers determined to be responsive thereto. For example, antibodies that block the interaction between PD-L1, PD-L2, and/or CTLA-4 and their receptors {e.g., PD-L1 binding to PD-1, PD-L2 binding to PD-1, and the like) can be used to treat cancer in subjects identified as likely responding thereto.

6. Pharmaceutical Compositions

In another aspect, the present invention provides pharmaceutically acceptable compositions which comprise a therapeutically-effective amount of an agent that modulates {e.g., decreases) biomarker expression and/or activity, formulated together with one or more pharmaceutically acceptable carriers (additives) and/or diluents. As described in detail below, the pharmaceutical compositions of the present invention may be specially formulated for administration in solid or liquid form, including those adapted for the following: (1) oral administration, for example, drenches (aqueous or non-aqueous solutions or suspensions), tablets, boluses, powders, granules, pastes; (2) parenteral administration, for example, by subcutaneous, intramuscular or intravenous injection as, for example, a sterile solution or suspension; (3) topical application, for example, as a cream, ointment or spray applied to the skin; (4) intravaginally or intrarectally, for example, as a pessary, cream or foam; or (5) aerosol, for example, as an aqueous aerosol, liposomal preparation or solid particles containing the compound.

The phrase "therapeutically-effective amount" as used herein means that amount of an agent that modulates (e.g., inhibits) biomarker expression and/or activity, or expression and/or activity of the complex, or composition comprising an agent that modulates (e.g.,

- I ll - inhibits) biomarker expression and/or activity, or expression and/or activity of the complex, which is effective for producing some desired therapeutic effect, e.g., cancer treatment, at a reasonable benefit/risk ratio.

The phrase "pharmaceutically acceptable" is employed herein to refer to those agents, materials, compositions, and/or dosage forms which are, within the scope of sound medical judgment, suitable for use in contact with the tissues of human beings and animals without excessive toxicity, irritation, allergic response, or other problem or complication, commensurate with a reasonable benefit/risk ratio.

The phrase "pharmaceutically-acceptable carrier" as used herein means a pharmaceutically-acceptable material, composition or vehicle, such as a liquid or solid filler, diluent, excipient, solvent or encapsulating material, involved in carrying or transporting the subject chemical from one organ, or portion of the body, to another organ, or portion of the body. Each carrier must be "acceptable" in the sense of being compatible with the other ingredients of the formulation and not injurious to the subject. Some examples of materials which can serve as pharmaceutically-acceptable carriers include: (1) sugars, such as lactose, glucose and sucrose; (2) starches, such as corn starch and potato starch; (3) cellulose, and its derivatives, such as sodium carboxymethyl cellulose, ethyl cellulose and cellulose acetate; (4) powdered tragacanth; (5) malt; (6) gelatin; (7) talc; (8) excipients, such as cocoa butter and suppository waxes; (9) oils, such as peanut oil, cottonseed oil, safflower oil, sesame oil, olive oil, corn oil and soybean oil; (10) glycols, such as propylene glycol; (11) polyols, such as glycerin, sorbitol, mannitol and

polyethylene glycol; (12) esters, such as ethyl oleate and ethyl laurate; (13) agar; (14) buffering agents, such as magnesium hydroxide and aluminum hydroxide; (15) alginic acid; (16) pyrogen-free water; (17) isotonic saline; (18) Ringer's solution; (19) ethyl alcohol; (20) phosphate buffer solutions; and (21) other non-toxic compatible substances employed in pharmaceutical formulations.

The term "pharmaceutically-acceptable salts" refers to the relatively non-toxic, inorganic and organic acid addition salts of the agents that modulates (e.g., inhibits) biomarker expression and/or activity, or expression and/or activity of the complex encompassed by the present invention. These salts can be prepared in situ during the final isolation and purification of the respiration uncoupling agents, or by separately reacting a purified respiration uncoupling agent in its free base form with a suitable organic or inorganic acid, and isolating the salt thus formed. Representative salts include the hydrobromide, hydrochloride, sulfate, bisulfate, phosphate, nitrate, acetate, valerate, oleate, palmitate, stearate, laurate, benzoate, lactate, phosphate, tosylate, citrate, maleate, fumarate, succinate, tartrate, napthylate, mesylate, glucoheptonate, lactobionate, and laurylsulphonate salts and the like (See, for example, Berge et al. (1977) "Pharmaceutical Salts", J. Pharm. Sci. 66: 1-19).

In other cases, the agents useful in the methods of the present invention may contain one or more acidic functional groups and, thus, are capable of forming pharmaceutically- acceptable salts with pharmaceutically-acceptable bases. The term "pharmaceutically- acceptable salts" in these instances refers to the relatively non-toxic, inorganic and organic base addition salts of agents that modulates (e.g., inhibits) biomarker expression and/or activity, or expression and/or activity of the complex. These salts can likewise be prepared in situ during the final isolation and purification of the respiration uncoupling agents, or by separately reacting the purified respiration uncoupling agent in its free acid form with a suitable base, such as the hydroxide, carbonate or bicarbonate of a pharmaceutically- acceptable metal cation, with ammonia, or with a pharmaceutically-acceptable organic primary, secondary or tertiary amine. Representative alkali or alkaline earth salts include the lithium, sodium, potassium, calcium, magnesium, and aluminum salts and the like. Representative organic amines useful for the formation of base addition salts include ethylamine, diethylamine, ethylenediamine, ethanolamine, diethanolamine, piperazine and the like (see, for example, Berge et al., supra).

Wetting agents, emulsifiers and lubricants, such as sodium lauryl sulfate and magnesium stearate, as well as coloring agents, release agents, coating agents, sweetening, flavoring and perfuming agents, preservatives and antioxidants can also be present in the compositions.

Examples of pharmaceutically-acceptable antioxidants include: (1) water soluble antioxidants, such as ascorbic acid, cysteine hydrochloride, sodium bisulfate, sodium metabi sulfite, sodium sulfite and the like; (2) oil-soluble antioxidants, such as ascorbyl palmitate, butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), lecithin, propyl gallate, alpha-tocopherol, and the like; and (3) metal chelating agents, such as citric acid, ethylenediamine tetraacetic acid (EDTA), sorbitol, tartaric acid, phosphoric acid, and the like.

Formulations useful in the methods of the present invention include those suitable for oral, nasal, topical (including buccal and sublingual), rectal, vaginal, aerosol and/or parenteral administration. The formulations may conveniently be presented in unit dosage form and may be prepared by any methods well known in the art of pharmacy. The amount of active ingredient which can be combined with a carrier material to produce a single dosage form will vary depending upon the host being treated, the particular mode of administration. The amount of active ingredient, which can be combined with a carrier material to produce a single dosage form will generally be that amount of the compound which produces a therapeutic effect. Generally, out of one hundred per cent, this amount will range from about 1 per cent to about ninety-nine percent of active ingredient, preferably from about 5 per cent to about 70 per cent, most preferably from about 10 per cent to about 30 per cent.

Methods of preparing these formulations or compositions include the step of bringing into association an agent that modulates (e.g., inhibits) biomarker expression and/or activity, with the carrier and, optionally, one or more accessory ingredients. In general, the formulations are prepared by uniformly and intimately bringing into association a respiration uncoupling agent with liquid carriers, or finely divided solid carriers, or both, and then, if necessary, shaping the product.

Formulations suitable for oral administration may be in the form of capsules, cachets, pills, tablets, lozenges (using a flavored basis, usually sucrose and acacia or tragacanth), powders, granules, or as a solution or a suspension in an aqueous or non- aqueous liquid, or as an oil-in-water or water-in-oil liquid emulsion, or as an elixir or syrup, or as pastilles (using an inert base, such as gelatin and glycerin, or sucrose and acacia) and/or as mouth washes and the like, each containing a predetermined amount of a respiration uncoupling agent as an active ingredient. A compound may also be administered as a bolus, electuary or paste.

In solid dosage forms for oral administration (capsules, tablets, pills, dragees, powders, granules and the like), the active ingredient is mixed with one or more

pharmaceutically-acceptable carriers, such as sodium citrate or dicalcium phosphate, and/or any of the following: (1) fillers or extenders, such as starches, lactose, sucrose, glucose, mannitol, and/or silicic acid; (2) binders, such as, for example, carboxymethylcellulose, alginates, gelatin, polyvinyl pyrrolidone, sucrose and/or acacia; (3) humectants, such as glycerol; (4) disintegrating agents, such as agar-agar, calcium carbonate, potato or tapioca starch, alginic acid, certain silicates, and sodium carbonate; (5) solution retarding agents, such as paraffin; (6) absorption accelerators, such as quaternary ammonium compounds; (7) wetting agents, such as, for example, acetyl alcohol and glycerol monostearate; (8) absorbents, such as kaolin and bentonite clay; (9) lubricants, such a talc, calcium stearate, magnesium stearate, solid polyethylene glycols, sodium lauryl sulfate, and mixtures thereof; and (10) coloring agents. In the case of capsules, tablets and pills, the

pharmaceutical compositions may also comprise buffering agents. Solid compositions of a similar type may also be employed as fillers in soft and hard-filled gelatin capsules using such excipients as lactose or milk sugars, as well as high molecular weight polyethylene glycols and the like.

A tablet may be made by compression or molding, optionally with one or more accessory ingredients. Compressed tablets may be prepared using binder (for example, gelatin or hydroxypropylmethyl cellulose), lubricant, inert diluent, preservative, disintegrant (for example, sodium starch glycolate or cross-linked sodium carboxymethyl cellulose), surface-active or dispersing agent. Molded tablets may be made by molding in a suitable machine a mixture of the powdered peptide or peptidomimetic moistened with an inert liquid diluent.

Tablets, and other solid dosage forms, such as dragees, capsules, pills and granules, may optionally be scored or prepared with coatings and shells, such as enteric coatings and other coatings well known in the pharmaceutical-formulating art. They may also be formulated so as to provide slow or controlled release of the active ingredient therein using, for example, hydroxypropylmethyl cellulose in varying proportions to provide the desired release profile, other polymer matrices, liposomes and/or microspheres. They may be sterilized by, for example, filtration through a bacteria-retaining filter, or by incorporating sterilizing agents in the form of sterile solid compositions, which can be dissolved in sterile water, or some other sterile injectable medium immediately before use. These compositions may also optionally contain opacifying agents and may be of a composition that they release the active ingredient(s) only, or preferentially, in a certain portion of the

gastrointestinal tract, optionally, in a delayed manner. Examples of embedding

compositions, which can be used include polymeric substances and waxes. The active ingredient can also be in micro-encapsulated form, if appropriate, with one or more of the above-described excipients.

Liquid dosage forms for oral administration include pharmaceutically acceptable emulsions, microemulsions, solutions, suspensions, syrups and elixirs. In addition to the active ingredient, the liquid dosage forms may contain inert diluents commonly used in the art, such as, for example, water or other solvents, solubilizing agents and emulsifiers, such as ethyl alcohol, isopropyl alcohol, ethyl carbonate, ethyl acetate, benzyl alcohol, benzyl benzoate, propylene glycol, 1,3-butylene glycol, oils (in particular, cottonseed, groundnut, corn, germ, olive, castor and sesame oils), glycerol, tetrahydrofuryl alcohol, polyethylene glycols and fatty acid esters of sorbitan, and mixtures thereof.

Besides inert diluents, the oral compositions can also include adjuvants such as wetting agents, emulsifying and suspending agents, sweetening, flavoring, coloring, perfuming and preservative agents.

Suspensions, in addition to the active agent may contain suspending agents as, for example, ethoxylated isostearyl alcohols, poly oxy ethylene sorbitol and sorbitan esters, microcrystalline cellulose, aluminum metahydroxide, bentonite, agar-agar and tragacanth, and mixtures thereof.

Formulations for rectal or vaginal administration may be presented as a suppository, which may be prepared by mixing one or more respiration uncoupling agents with one or more suitable nonirritating excipients or carriers comprising, for example, cocoa butter, polyethylene glycol, a suppository wax or a salicylate, and which is solid at room temperature, but liquid at body temperature and, therefore, will melt in the rectum or vaginal cavity and release the active agent.

Formulations which are suitable for vaginal administration also include pessaries, tampons, creams, gels, pastes, foams or spray formulations containing such carriers as are known in the art to be appropriate.

Dosage forms for the topical or transdermal administration of an agent that modulates (e.g., inhibits) biomarker expression and/or activity include powders, sprays, ointments, pastes, creams, lotions, gels, solutions, patches and inhalants. The active component may be mixed under sterile conditions with a pharmaceutically-acceptable carrier, and with any preservatives, buffers, or propellants which may be required.

The ointments, pastes, creams and gels may contain, in addition to a respiration uncoupling agent, excipients, such as animal and vegetable fats, oils, waxes, paraffins, starch, tragacanth, cellulose derivatives, polyethylene glycols, silicones, bentonites, silicic acid, talc and zinc oxide, or mixtures thereof.

Powders and sprays can contain, in addition to an agent that modulates (e.g., inhibits) biomarker expression and/or activity, excipients such as lactose, talc, silicic acid, aluminum hydroxide, calcium silicates and polyamide powder, or mixtures of these substances. Sprays can additionally contain customary propellants, such as chlorofluorohydrocarbons and volatile unsubstituted hydrocarbons, such as butane and propane.

The agent that modulates (e.g., inhibits) biomarker expression and/or activity, can be alternatively administered by aerosol. This is accomplished by preparing an aqueous aerosol, liposomal preparation or solid particles containing the compound. A nonaqueous (e.g., fluorocarbon propellant) suspension could be used. Sonic nebulizers are preferred because they minimize exposing the agent to shear, which can result in degradation of the compound.

Ordinarily, an aqueous aerosol is made by formulating an aqueous solution or suspension of the agent together with conventional pharmaceutically acceptable carriers and stabilizers. The carriers and stabilizers vary with the requirements of the particular compound, but typically include nonionic surfactants (Tweens, Pluronics, or polyethylene glycol), innocuous proteins like serum albumin, sorbitan esters, oleic acid, lecithin, amino acids such as glycine, buffers, salts, sugars or sugar alcohols. Aerosols generally are prepared from isotonic solutions.

Transdermal patches have the added advantage of providing controlled delivery of a respiration uncoupling agent to the body. Such dosage forms can be made by dissolving or dispersing the agent in the proper medium. Absorption enhancers can also be used to increase the flux of the peptidomimetic across the skin. The rate of such flux can be controlled by either providing a rate controlling membrane or dispersing the

peptidomimetic in a polymer matrix or gel.

Ophthalmic formulations, eye ointments, powders, solutions and the like, are also contemplated as being within the scope of this invention.

Pharmaceutical compositions of this invention suitable for parenteral administration comprise one or more respiration uncoupling agents in combination with one or more pharmaceutically-acceptable sterile isotonic aqueous or nonaqueous solutions, dispersions, suspensions or emulsions, or sterile powders which may be reconstituted into sterile injectable solutions or dispersions just prior to use, which may contain antioxidants, buffers, bacteriostats, solutes which render the formulation isotonic with the blood of the intended recipient or suspending or thickening agents.

Examples of suitable aqueous and nonaqueous carriers which may be employed in the pharmaceutical compositions of the present invention include water, ethanol, polyols (such as glycerol, propylene glycol, polyethylene glycol, and the like), and suitable mixtures thereof, vegetable oils, such as olive oil, and injectable organic esters, such as ethyl oleate. Proper fluidity can be maintained, for example, by the use of coating materials, such as lecithin, by the maintenance of the required particle size in the case of dispersions, and by the use of surfactants.

These compositions may also contain adjuvants such as preservatives, wetting agents, emulsifying agents and dispersing agents. Prevention of the action of

microorganisms may be ensured by the inclusion of various antibacterial and antifungal agents, for example, paraben, chlorobutanol, phenol sorbic acid, and the like. It may also be desirable to include isotonic agents, such as sugars, sodium chloride, and the like into the compositions. In addition, prolonged absorption of the injectable pharmaceutical form may be brought about by the inclusion of agents which delay absorption such as aluminum monostearate and gelatin.

In some cases, in order to prolong the effect of a drug, it is desirable to slow the absorption of the drug from subcutaneous or intramuscular injection. This may be accomplished by the use of a liquid suspension of crystalline or amorphous material having poor water solubility. The rate of absorption of the drug then depends upon its rate of dissolution, which, in turn, may depend upon crystal size and crystalline form.

Alternatively, delayed absorption of a parenterally-administered drug form is accomplished by dissolving or suspending the drug in an oil vehicle.

Injectable depot forms are made by forming microencapsule matrices of an agent that modulates (e.g., inhibits) biomarker expression and/or activity, in biodegradable polymers such as polylactide-polyglycolide. Depending on the ratio of drug to polymer, and the nature of the particular polymer employed, the rate of drug release can be controlled. Examples of other biodegradable polymers include poly(orthoesters) and poly(anhydrides). Depot injectable formulations are also prepared by entrapping the drug in liposomes or microemulsions, which are compatible with body tissue.

When the respiration uncoupling agents of the present invention are administered as pharmaceuticals, to humans and animals, they can be given per se or as a pharmaceutical composition containing, for example, 0.1 to 99.5% (more preferably, 0.5 to 90%) of active ingredient in combination with a pharmaceutically acceptable carrier.

Actual dosage levels of the active ingredients in the pharmaceutical compositions of this invention may be determined by the methods of the present invention so as to obtain an amount of the active ingredient, which is effective to achieve the desired therapeutic response for a particular subject, composition, and mode of administration, without being toxic to the subject.

The nucleic acid molecules of the present invention can be inserted into vectors and used as gene therapy vectors. Gene therapy vectors can be delivered to a subject by, for example, intravenous injection, local administration (see U.S. Pat. No. 5,328,470) or by stereotactic injection (see e.g., Chen et al. (1994) Proc. Natl. Acad. Sci. USA 91 :3054 3057). The pharmaceutical preparation of the gene therapy vector can include the gene therapy vector in an acceptable diluent, or can comprise a slow release matrix in which the gene delivery vehicle is imbedded. Alternatively, where the complete gene delivery vector can be produced intact from recombinant cells, e.g., retroviral vectors, the pharmaceutical preparation can include one or more cells which produce the gene delivery system.

The present invention also encompasses kits for detecting and/or modulating biomarkers described herein. A kit of the present invention may also include instructional materials disclosing or describing the use of the kit or an antibody of the disclosed invention in a method of the disclosed invention as provided herein. A kit may also include additional components to facilitate the particular application for which the kit is designed. For example, a kit may additionally contain means of detecting the label (e.g., enzyme substrates for enzymatic labels, filter sets to detect fluorescent labels, appropriate secondary labels such as a sheep anti-mouse-HRP, etc.) and reagents necessary for controls (e.g., control biological samples or standards). A kit may additionally include buffers and other reagents recognized for use in a method of the disclosed invention. Non-limiting examples include agents to reduce non-specific binding, such as a carrier protein or a detergent.

Exemplification

This invention is further illustrated by the following examples, which should not be construed as limiting.

Example 1: Materials and Methods for Example 2

a. Clinical cohort consolidation

The training cohort was gathered from patients enrolled in CA209-009

(NCTO 1358721), a study of nivolumab (BMS-936558) monotherapy in metastatic renal cell carcinoma. The validation cohort was gathered from patients at the Dana-Farber Cancer Institute and Memorial Sloan Kettering Cancer Institute who received immune checkpoint therapy as monotherapy or in combination with other immune checkpoint or targeted therapies and had banked adequate pre-treatment tumor tissue for whole exome

characterization. All patients provided consent to an Institutional Review Board protocol that allows research molecular characterization of tumor and germline samples.

b. DNA and RNA extraction and sequencing

After fixation and mounting, 5-10 ΙΟμπι slices from formalin-fixed, paraffin- embedded (FFPE) tumor blocks were obtained, and tumor-enriched tissue was

macrodissected. Paraffin was removed from FFPE sections and cores using CitriSolv™ (Fisher Scientific, Hampton, NH), followed by ethanol washes and tissue lysis overnight at 56°C. Samples were then incubated at 90°C to remove DNA crosslinks, and DNA- and when possible, RNA- extraction was performed using Qiagen AllPrep DNA/RNA Mini Kit (#51306, Qiagen, Hilden, Germany). Germline DNA was obtained from adjacent PBMCs.

Whole exome and whole transcriptome sequencing of tumor and germline samples were performed as previously described (Van Allen et al. (2015) Science 350:207-211; Van Allen et al. (2014) Nat. Med. 20:682-688). All samples in the training cohort were sequenced using the Illumina exome, while a portion of the samples in the validation cohort were sequenced using the Agilent exome (Table 4A). The Illumina exome uses Illumina' s in-solution DNA probe based hybrid selection method to target approximately 37.7Mb of mainly exonic territory, using similar principles as the Broad Institute- Agilent Technologies developed in-solution RNA probe based hybrid selection method (Agilent SureSelect ^™ All Exon V2) (Fisher et al. (2011) Genome Biol. 12:R1; Gnirke et al. (2009) Nat. Biotechnol. 27: 182-189) to generate Illumina exome sequencing libraries.

Pooled libraries were normalized to 2nM and denatured using 0.2 N NaOH prior to sequencing. Flowcell cluster amplification and sequencing were performed according to the manufacturer's protocols using either the HiSeq 2000 v3 or HiSeq 2500. Each run was a 76bp paired-end with a dual eight-base index barcode read. Data was analyzed using the Broad Picard Pipeline, which includes de-multiplexing and data aggregation.

Exome sequence data processing was performed using established analytical pipelines at the Broad Institute. A BAM file was produced using the Picard pipeline (at the World Wide Web address of picard.sourceforge.net), which aligns the tumor and normal sequences to the hgl9 human genome build using Illumina sequencing reads. The BAM was uploaded into the Firehose pipeline (at the World Wide Web address of

broadinstitute.org/cancer/cga/Firehose), which manages input and output files to be executed by GenePattern (Reich et al. (2006) Nat. Genet. 38:500-501). Samples with mean target coverage less than 25x in the tumor and less than 15x in matched normal were excluded.

Quality control modules within Firehose were applied to all sequencing data for comparison of the origin of tumor and normal genotypes and to assess fingerprinting concordance. Cross-contamination of samples was estimated using ContEst (Cibulskis et al. (2011) Bioinformatics 27:2601-2602). Samples with ContEst estimates exceeding 5% were excluded from analysis.

c. Whole exome and whole transcriptome analyses

MuTect was applied to identify somatic single-nucleotide variants (Cibulskis et al.

(2013) Nat. Biotechnol. 31 :213-219). Strelka was used to identify somatic insertions and deletions (Saunders et al. (2012) Bioinformatics 28: 1811-1817) across the whole exome. Indelocator, which detects small insertions and deletions after local realignment of tumor and normal sequences, was additionally applied to provide further sensitivity to detect indels in PBRMl (Cancer Genome Atlas Research (2011) Nature 474:609-615). The union of indels called by Strelka and Indelocator was used for final analysis. Artifacts introduced by DNA oxidation during sequencing were computationally removed using a filter-based method (Costello et al. (2013) Nuc. Acids Res. 41 :e67). All somatic mutations detected by whole-exome sequencing were analyzed for potential false positive calls by performing a comparison to mutation calls from a panel of 2,500 germline DNA samples (Stachler et al. (2015) Nat. Genet. 47: 1047-1055). Mutations found in germline samples were removed from analysis. Annotation of identified variants was done using Oncotator (available at the World Wide Web address of www.broadinstitute.org/cancer/cga/oncotator). All nonsynonymous alterations in PBRMl were manually reviewed in Integrated Genomics Viewer (IGV 2.3.57) for sequencing quality (Thorvaldsdottir et al. (2013) Brief Bioinform 14: 178-192).

Copy ratios were calculated for each captured target by dividing the tumor coverage by the median coverage obtained in a set of reference normal samples. The resulting copy ratios were segmented using the circular binary segmentation algorithm (Olshen et al. (2004) Biostatistics 5:557-572). Allelic copy number alterations were called while taking into account sample-specific overall chromosomal aberrations (focality) (Brastianos et al. (2015) Cancer Discov. 5: 1164-1177). Inference of mutational clonality, tumor purity, and tumor ploidy was accomplished with ABSOLUTE (Carter et al. (2012) Nat Biotechnol. 30:413-421). Samples had to have estimated tumor purity greater than 10% to be included in the final analysis. As a final quality control metric to ensure adequate sequencing coverage and tumor purity to detect relevant oncogenic events, all samples had to have at least one nonsynonymous mutation in at least one high confidence or candidate cancer driver gene to be included in the final analysis (Tamborero et al. (2013) Sci. Rep. 3 :2650).

The 4-digit HLA type for each sample was inferred using Polysolver (Shukla et al. (2015) Nat. Biotechnol. 33 : 1152-1158). Neo-epitopes were predicted for each patient by defining all novel amino acid 9mers and lOmers resulting from each single nucleotide polymorphism and indel and determining whether the predicted binding affinity to the patient's germline HLA alleles was <500 nM using NetMHCpan (v2.4) (Hoof et al. (2009) Immunogenetics 61 : 1-13; Karosiene et al. (2013) Immunogenetics 65:711-724; Nielsen et al. (2007) PLoS One 2:e796).

d. TCGA Analysis

Whole exome mutations annotation files (MAFs) and whole transcriptome gene expression data (RSEM) were downloaded from the Firebrowse KIRC TCGA data release (2016 01 28). Samples with whole transcriptome sequencing in normal tissue only, as well as samples derived from FFPE (N=3), were excluded from analysis.

e. Serum biomarker analyses

Serum biomarker analyses were performed as described previously in Choueiri et al. (2016) Clin. Cancer Res. 22:5461-5471.

f . Statistical analyses

Comparisons of neoantigen and mutational load between response groups and of expression levels of individual genes between PBRMl -truncated and PBRMl-wildtype tumors were done with the non-parametric Wilcoxon rank-sum test. Comparisons of the proportion of patients with truncating alterations in PBRMl by clinical response group were done with the Pearson's chi-squared test without continuity correction. Kaplan-Meier analyses were done using the R packages survival and survminer. All comparisons were two-sided with an alpha-level of 0.05. All statistical analyses were done in R version 3.2.3.

Given the low mutational burden and high tumor microenvironment immune activity characteristic of renal cell carcinoma, it is believed that specific somatic genetic features other than mutational load drive response to immune checkpoint inhibitors in ccRCC. As part of a prospective clinical trial (Choueiri et al. (2016), supra), Applicants analyzed a clinical cohort of 91 patients with metastatic clear cell renal cell carcinoma (mRCC) treated with anti-PDl therapy (nivolumab) (Figure 1A; Arm 1 : 0.3 mg/kg (N=22); Arm 2: 2 mg/kg (N=22); Arm 3 : 10 mg/kg (N=23), and Arm 4: 10 mg/kg (N=24)). Among 56 patients with attempted whole exome sequenceing of pre-treatment tumors, 34 had high- quality whole exome sequencing (WES) for discovery of genetic predictors of response to immune checkpoint therapy, and then validated the findings in an independent cohort of WES of pre-treatment tumors from 28 patients (Figure IB). Applicants also analyzed pre- treatment whole transcriptome sequencing (WTS) from a subset of 42 patients from both the training and validation cohorts to assess the impact of genetic changes associated with treatment response on tumor gene expression and immune infiltration.

Example 2: Loss-of-function of PBRM1 correlates with response to anti-PDl PD-Ll therapy in renal cell carcinoma

Quality-control metrics were applied to both the training and validation cohorts to ensure sensitive mutation detection (Cibulskis et al. (2011), supra) (Figure IB and Table 2A). Of the samples included in the final analysis, average exome-wide target coverage was 140-fold for tumor samples (range: 27-210) and 91-fold (range: 48-168) for matched germline samples. Analysis methods used herein include somatic mutation identification (single nucleotide polymorphisms and insertions and deletions) (as in Cibulskis et al.

(2013), supra and Saunders et al. (2012) Bioinformatics 28: 1811-1817), human lymphocyte antigen (ELLA) typing from germline WES (as in Shukla et al. (2015) Nat. Biotechnol.

33 : 1152-1158), neoantigen prediction (as in Hoof et al. (2009) Immunogenetics 61 : 1-13), and estimation of mutational clonality and tumor purity and ploidy (Carter et al. (2012) Nat Biotechnol. 30:413-421) using established methods (as in Example 1 and Figure IB). In the training cohort, of the 56 out of 91 patients for whom adequate pre-treatment tissue was available for WES, 34 passed quality control and were included in the final analysis (Figure IB). For example, sample VA1008 having a chromosome 3p deletion was excluded as having low tumor purity (estimated tumor purity = 0.11). Among these 56 pairs matched tumor and normal samples, Sample 2 664 contains germline BAM only, while Sample 4 49 contains tumor BAM only. As quality control for sequenced tissure, 6 of 56 samples were excluded due to poor tumor coverage. They were Samples 4_54 (0.079 x), 9_47 (0.30 x), 8_100 (7.71 x), 11 5 (8.69 x), 1_72 (9.63 x), and 9_66 (8.72 x). Another sample, 9 119 (26.88 x), was not excluded. For this sample, with estimated tumor purity of 0.49 and mean target coverage of 27 x, a sensitivity of -90% detected a heterozygous mutation in CA- 209009-9 119 (see Cibulskis et al. (2013), supra). Quality control for copy number was also performed.

Table 2A: Sequencing Metrics and Inclusion/Exclusion Criteria for Whole Exome

Sequencing in Training Cohort (N=56)

CA209009 5 157.143939 89.39856 whole exome illu 0.2 3.6 1 0 1 mina coding vl

CA209009 5 176.007671 81.059438 whole exome illu 0.35 1.9 0 0 106 mina coding vl

CA209009 5 139.328276 75.654059 whole exome illu 0.21 2.3 0 0 18 mina coding vl

CA209009 5 178.624687 105.356301 whole exome illu 0.51 3.39 1 0 21 mina coding vl

CA209009 5 138.664874 93.93237 whole exome illu 0.19 4.28 1 0 41 mina coding vl

CA209009 5 162.205322 85.879444 whole exome illu 0.31 1.81 0 0 50 mina coding vl

CA209009 5 158.127987 100.10628 whole exome illu 0.6 1.83 0 0

_73 mina coding vl

CA209009 6 147.571574 114.169462 whole exome illu 0.13 1.92 0 0 39 mina coding vl

CA209009 8 152.057615 91.424807 whole exome illu 0.48 2.06 0 0 105 mina coding vl

CA209009 9 26.875509 90.734659 whole exome illu 0.49 3.08 1 0 119 mina coding vl

CA209009 9 125.149722 97.245404 whole exome illu 0.34 1.93 0 0

_27 mina coding vl

CA209009 9 131.064027 90.415506 whole exome illu 0.54 1.88 0 0 _52 mina coding vl

CA209009 9 210.012354 98.486524 whole exome illu 0.38 2.2 0 0

_97 mina coding vl

CA209009 2 43.586957 168.436641 whole exome illu 0.13 4.12 1 0 85 mina coding vl

CA209009 5 159.912441 69.844188 whole exome illu 0.52 1.68 0 EarlyDe 2 mina coding vl ath

CA209009 5 150.205436 89.123637 whole exome illu NA NA NA EarlyDe 29 mina coding vl ath

CA209009 6 34.101887 117.822339 whole exome illu 0.36 2.77 1 EarlyDe 99 mina coding vl ath

CA209009 1 9.627872 94.01896 whole exome illu NA NA NA LowCov

_72 mina coding vl erage

CA209009 1 8.689284 89.713424 whole exome illu 0.36 1.98 0 LowCov 1 5 mina coding vl erage

CA209009 4 0.007939 84.883698 whole exome illu NA NA NA LowCov _54 mina coding vl erage

CA209009 8 7.711684 105.962605 whole exome illu 0.34 2.01 0 LowCov 100 mina coding vl erage

CA209009 9 0.298156 95.4427 whole exome illu NA NA NA LowCov

_47 mina coding vl erage

CA209009 9 8.71954 98.033649 whole exome illu 0.46 2.16 0 LowCov 66 mina coding vl erage

CA209009 1 105.603458 72.354112 whole exome illu 0.06 2.43 0 LowPuri

_43 mina coding vl t CA209009 1 162.560923 104.266666 whole exome illu 0.05 2.74 0 LowPuri 1 12 mina coding vl t

CA209009 1 166.047506 75.247762 whole exome illu 0.1 2.46 0 LowPuri 1 24 mina coding vl ty

CA209009 1 154.736269 87.045058 whole exome illu 0.1 2.44 0 LowPuri 1 40 mina coding vl ty

CA209009 1 154.801856 83.048353 whole exome illu NA NA NA LowPuri 1 8 mina coding vl ty

CA209009 1 138.626523 96.365324 whole exome illu NA NA NA LowPuri 3 103 mina coding vl ty

CA209009 3 159.566974 100.887491 whole exome illu 0.07 2.96 0 LowPuri 26 mina coding vl ty

CA209009 4 143.956046 90.060356 whole exome illu 0.09 2.57 0 LowPuri 95 mina coding vl ty

CA209009 5 129.343681 81.980679 whole exome illu 0.04 3.61 1 LowPuri

_n mina coding vl ty

CA209009 5 144.076612 97.672268 whole exome illu 0.06 2.91 0 LowPuri

_22 mina coding vl ty

CA209009 5 162.443009 89.968028 whole exome illu 0.08 2.45 0 LowPuri 28 mina coding vl ty

CA209009 5 145.806274 83.646769 whole exome illu 0.07 2.69 0 LowPuri

_6 mina coding vl ty

CA209009 9 132.158193 79.179771 whole exome illu 0.06 2.58 0 LowPuri _45 mina coding vl ty

To stratify clinical cohort between patients who most clearly derived durable clinical benefit from anti-PDl therapy and those who did not, three response categories were defined based on a composite end point incorporating RECIST criteria (Eisenhauer et al. (2009) Eur. J. Cancer 45:228-247), tumor shrinkage, and progression-free survival (PFS) (Figures 2-3 and Table 2B). "Extreme responders" included all patients with complete response (CR) or partial response (PR) by RECIST. Patients with stable disease (SD) as their best response by RECIST were also considered extreme responders if they had objective reduction in tumor size lasting at least 6 months, such as at least 12 months.

"Extreme progressors" experienced early tumor growth: progressive disease (PD) by

RECIST as best response with progression in less than 3 months. An intermediate group of patients who experienced SD or PR with objective tumor shrinkage lasting less than 6 months (or sometimes less than 12 months as indicated in certain figures) or PD with PFS longer than 3 months were called "intermediate benefit." One patient (5 50) was classified as an "extreme responder" despite experiencing a short period of early tumor progression (PFS = 2.9 months), which likely represented pseudo-progression, as further follow-up showed sustained tumor remission (Figures IB and 2A-2B). Three patients who experienced death on-treatment prior to the first staging scans were excluded from analysis (Table 2A). Not evaluable (NE): No RECIST evaluation made. Mixed response (X): Simultaneous tumor shrinkage and growth.

Detailed clinical information and immunohistochemical staining was available for all 91 patients in the training cohort. Pre-treatment immunohistochemical staining for PD- Ll was positive at >1% for 30% of patients and at >5% for 16% of patients, which is generally representative of other large cohorts of clear cell RCC (Table 2B) (Motzer et al. (2015), supra). Duration of overall survival did not vary significantly by dose of therapy, patient gender, or PD-L1 immunohistochemical staining in Kaplan-Meier analyses, while objective tumor response by RECIST substantially prolonged duration of overall survival (Figure 3A-3D). For a summary of results, see Figure 11.

Table 2B: Clinical characteristics of patients receiving anti-PDl therapy (nivolumab) in training cohort (N=91)

CA20 0 M 8 10 SD 2 817 1 529 0 100 0 0 0 0 0 extrem

9009 2 mg/kg e respo 14 10 nder

7

CA20 0 F 6 2 PR -59 957 1 500 0 95 2 2 1 1 0 extrem

9009 1 mg kg e respo 9 34 nder

CA20 1 M 6 10 PR -43 684 0 500 0 100 0 0 0 0 0 extrem

9009 4 mg kg e respo 11 93 nder

CA20 0 M 6 0.3 SD -44 1003 1 499 0 100 0 0 0 0 0 extrem

9009 9 mg/kg e respo 9 47 nder

CA20 0 F 4 10 SD -31 912 1 463 0 100 0 0 0 0 0 extrem

9009 1 mg/kg e respo 15 94 nder

CA20 1 M 7 10 PR -52 773 1 414 0 100 0 0 0 0 0 extrem

9009 2 mg/kg e respo 9 119 nder

CA20 0 M 6 10 SD -3 834 1 374 0 100 0 0 0 0 0 extrem

9009 8 mg/kg- e respo 13 11 N nder 1

CA20 0 F 6 10 SD 0 1094 1 337 0 100 0 0 0 0 0 extrem

9009 4 mg/kg- e respo 11 13 N nder

CA20 0 F 7 2 SD -10 821 0 295 1 100 0 0 0 0 0 stable_

9009 0 mg/kg disease

15 75

CA20 0 M 4 10 SD 0 969 1 295 0 95 4 1 0 1 1 stable_

9009 8 mg/kg disease

11 57

CA20 0 M 6 0.3 PR -73 1051 1 292 0 97 3 0 0 1 0 extrem

9009 0 mg/kg e respo 11 8 nder

CA20 0 M 5 10 SD 4 862 1 289 0 94 3 2 1 1 0 stable_

9009 1 mg/kg disease

13 10

3

CA20 0 F 6 0.3 SD 8 976 1 254 0 100 0 0 0 0 0 stable_

9009 5 mg/kg disease

4 54

CA20 0 F 6 10 PR -37 365 0 246 0 100 0 0 0 0 0 extrem

9009 0 mg/kg e respo 9 30 nder

CA20 1 F 6 10 SD -5 995 1 246 0 100 0 0 0 0 0 stable_

9009 3 mg/kg- disease

9 52 N

CA20 0 M 7 2 SD -21 293 0 237 0 100 0 0 0 0 0 stable_

9009 8 mg/kg disease

14 89

CA20 0 M 6 2 SD 5 914 1 220 0 100 0 0 0 0 0 stable_

9009 5 mg/kg disease

5 4

CA20 1 M 6 10 SD -13 240 0 213 1 100 0 0 0 0 0 stable_

9009 5 mg/kg- disease

1 32 N

CA20 0 M 4 10 PR -44 662 0 209 0 99 1 0 0 1 0 extrem

9009 2 mg/kg e respo 15 76 nder

CA20 1 F 5 0.3 PR -51 340 1 208 0 100 0 0 0 0 0 extrem

9009 7 mg/kg e respo 3 114 nder

CA20 0 M 6 2 PR -43 197 0 197 0 70 5 10 15 1 1 extrem

9009 3 mg/kg e respo 5 22 nder

CA20 0 F 5 2 SD 5 798 1 184 1 100 0 0 0 0 0 stable_

9009 8 mg/kg disease

8 100

CA20 1 F 6 10 SD 0 1058 1 173 1 97 3 0 0 1 0 stable_

9009 4 mg/kg- disease

11 10 N

CA20 0 M 5 10 SD 17 772 0 171 0 98 2 0 0 1 0 stable_

9009 4 mg/kg disease 10 11

2

CA20 0 M 5 10 SD -4 169 1 169 0 100 0 0 0 0 0 stable_

9009 5 mg/kg- disease

5 17 N

CA20 0 M 6 0.3 SD 4 440 0 163 0 100 0 0 0 0 0 stable_

9009 7 mg kg disease

9 74

CA20 0 M 5 10 SD 2 766 0 157 1 100 0 0 0 0 0 stable_

9009 9 mg kg disease

2 64

CA20 1 F 6 2 SD 9 873 1 130 0 100 0 0 0 0 0 stable_

9009 1 mg kg disease

11 79

CA20 0 M 6 10 SD -2 149 1 127 1 100 0 0 0 0 0 stable_

9009 6 mg/kg- disease

5 23 N

CA20 0 M 6 10 SD 3 605 0 123 0 99 1 0 0 0 0 stable_

9009 0 mg/kg disease

4 49

CA20 0 M 5 0.3 SD 12 954 1 123 0 100 0 0 0 0 0 stable_

9009 7 mg/kg disease

11 71

CA20 1 M 5 10 SD 9 1024 1 122 0 100 0 0 0 0 0 stable_

9009 0 mg/kg- disease

11 11 N

CA20 1 M 6 0.3 SD 3 165 1 108 0 100 0 0 0 0 0 stable_

9009 4 mg/kg disease

2 102

CA20 0 F 5 10 SD -19 155 0 99 0 NA NA NA NA NA N stable_

9009 9 mg/kg- A disease

5 6 N

CA20 1 F 5 0.3 SD 8 680 0 88 0 95 5 0 0 0 0 stable_

9009 5 mg/kg disease

2 84

CA20 1 F 4 10 SD 7 106 1 87 0 NA NA NA NA NA N stable_

9009 8 mg/kg A disease

1 62

CA20 1 M 6 2 SD 15 366 1 87 0 100 0 0 0 0 0 stable_

9009 0 mg/kg disease

12 11

5

CA20 0 M 8 0.3 PD 29 177 0 86 0 98 1 1 0 1 0 extrem

9009 2 mg/kg e_progr

1 118 essor

CA20 1 F 6 10 SD -67 982 1 86 0 91 5 3 1 1 1 extrem

9009 3 mg/kg- e respo

5 50 N nder

CA20 0 F 6 10 SD 0 492 0 85 0 100 0 0 0 0 0 stable_

9009 3 mg/kg- disease

11 5 N

CA20 0 M 7 10 SD 7 464 0 85 0 100 0 0 0 0 0 stable_

9009 1 mg/kg disease

1 86

CA20 0 M 6 0.3 SD 4 147 0 82 0 100 0 0 0 0 0 stable_

9009 1 mg/kg disease

14 80

CA20 0 M 4 0.3 NE 81 0 81 0 95 3 2 0 0 1 not ev

9009 2 mg/kg aluable

2 42

CA20 0 M 6 0.3 SD 8 991 1 81 0 84 10 5 1 1 1 stable_

9009 4 mg/kg disease

14 59

CA20 1 F 6 2 SD 17 992 1 81 0 100 0 0 0 0 0 stable_

9009 2 mg/kg disease

11 56

CA20 0 M 3 10 PD 8 974 1 80 0 100 0 0 0 0 0 extrem

9009 2 mg/kg- e_progr

11 40 N essor

CA20 0 F 6 2 SD 19 350 1 75 0 38 26 30 6 1 1 stable_

9009 2 mg/kg disease

14 87 CA20 1 M 7 2 PD 11 178 0 58 0 100 0 0 0 0 0 extrem

9009 7 mg kg e_progr

5 73 essor

CA20 0 M 7 10 PD 6 62 0 51 0 100 0 0 0 0 0 extrem

9009 0 mg/kg- e_progr

11 24 N essor

CA20 1 M 6 10 PD 17 118 0 45 0 100 0 0 0 0 0 extrem

9009 4 mg kg e_progr

8 105 essor

CA20 0 M 4 0.3 PD 24 486 0 45 0 100 0 0 0 0 0 extrem

9009 6 mg kg e_progr

15 83 essor

CA20 0 M 7 10 PD -8 968 1 44 0 100 0 0 0 0 0 extrem

9009 4 mg/kg- e_progr

1 43 N essor

CA20 0 M 5 0.3 PD 1 132 0 44 0 94 6 0 0 1 0 extrem

9009 0 mg/kg e_progr

15 77 essor

CA20 0 M 5 10 PD 6 283 0 43 0 95 5 0 0 1 1 extrem

9009 8 mg/kg e_progr

1 72 essor

CA20 0 M 4 10 PD 30 43 1 43 0 97 1 2 0 1 0 extrem

9009 7 mg/kg- e_progr

5 28 N essor

CA20 1 M 6 0.3 PD 65 111 1 43 0 NA NA NA NA NA N extrem

9009 8 mg/kg A e_progr

5 18 essor

CA20 1 M 6 10 SD -13 349 1 43 1 96 4 0 0 1 0 stable_

9009 4 mg/kg disease

5 21

CA20 1 M 6 10 PD 10 195 1 42 0 100 0 0 0 0 0 extrem

9009 6 mg/kg e_progr

5 41 essor

CA20 1 M 3 2 SD -12 712 0 42 1 95 2 2 1 1 0 stable_

9009 6 mg/kg disease

2 85

CA20 0 M 5 2 NE 41 0 41 0 97 3 0 0 1 0 not ev

9009 0 mg/kg aluable

6 99

CA20 1 M 6 0.3 PD -4 740 0 39 0 100 0 0 0 0 0 extrem

9009 9 mg/kg e_progr

4 68 essor

CA20 0 M 6 10 PD 0 955 1 39 0 95 5 0 0 1 1 extrem

9009 9 mg/kg- e_progr

11 38 N essor

CA20 0 M 4 10 PD 12 967 1 39 0 100 0 0 0 0 0 extrem

9009 5 mg/kg- e_progr

13 51 N essor

CA20 1 F 7 10 PD 16 167 0 39 0 100 0 0 0 0 0 extrem

9009 2 mg/kg- e_progr

11 25 N essor

CA20 1 M 4 10 PD 24 43 1 38 0 100 0 0 0 0 0 extrem

9009 9 mg/kg e_progr

5 1 essor

CA20 1 M 5 10 SD -10 64 1 38 1 98 2 0 0 0 0 stable_

9009 8 mg/kg- disease

3 117 N

CA20 1 F 7 2 PD 12 582 1 37 0 100 0 0 0 0 0 extrem

9009 5 mg/kg e_progr

13 96 essor

CA20 1 M 7 0.3 PD 13 306 0 37 0 100 0 0 0 0 0 extrem

9009 2 mg/kg e_progr

6 39 essor

CA20 1 M 5 2 PD 22 871 1 37 0 100 0 0 0 0 0 extrem

9009 4 mg/kg e_progr

9 97 essor

CA20 0 F 4 10 PD 53 303 0 37 0 100 0 0 0 0 0 extrem

9009 0 mg/kg e_progr

14 69 essor

CA20 0 F 5 10 PD 64 43 1 37 0 97 2 1 0 1 0 extrem

9009 7 mg/kg e_progr

5 3 essor CA20 0 M 4 2 PD 38 470 0 36 0 100 0 0 0 0 0 extrem 9009 7 mg kg e_progr 9 66 essor

CA20 0 M 5 2 PD 47 92 0 36 0 85 10 4 1 1 1 extrem 9009 5 mg kg e_progr 4 95 essor

CA20 1 M 5 0.3 PD 57 499 0 36 0 25 35 20 20 1 1 extrem 9009 4 mg kg e_progr 13 90 essor

CA20 1 M 5 0.3 PD 36 968 1 35 0 68 30 2 0 1 1 extrem 9009 9 mg/kg e_progr 9 27 essor

CA20 0 M 6 0.3 PD -49 398 0 32 0 97 1 1 1 1 0 extrem 9009 5 mg/kg e_progr 14 98 essor

CA20 0 M 4 0.3 NE 31 0 31 0 100 0 0 0 0 0 not ev 9009 8 mg/kg aluable 5 2

CA20 0 M 5 2 NE 33 0 28 0 91 5 3 1 1 1 not ev 9009 5 mg/kg aluable 3 26

CA20 0 F 7 2 NE 189 1 22 0 95 3 1 1 1 1 not ev 9009 1 mg/kg aluable 5 63

CA20 0 M 6 0.3 NE 937 1 1 1 100 0 0 0 0 0 not ev 9009 7 mg/kg aluable 5 29

CA20 0 F 6 10 NE 741 0 1 1 NA NA NA NA NA N not ev 9009 9 mg/kg- A aluable 13 36 N

CA20 0 F 6 10 NE 179 0 1 1 NA NA NA NA NA N not ev 9009 9 mg/kg A aluable 11 12

All patients listed in Table 2B were treated with nivolumab. For sex, M represents male and F represents female.

On whole exome analysis, overall mutational loads were moderate in the 34 patients with high-quality WES in the training cohort (median 116.5, range 70-255), and mutational burden did not predict response to therapy, nor did burden of clonal mutations or the ratio of subclonal to clonal mutations (p>0.05 for all; Wilcoxon rank sum) (Figure 4A-4B, Table 2C). Thus, the role of mutations in particular genes in mediating response needs to be solved. To identify significantly mutated genes in this cohort, MutSigCV (Lawrence et al. (2013) Nature 499:214-218, available at the website of the Broad Insitute of the World Wide Web address of

software. broadinstitute.org/cancer/software/genepattern/modules/docs/ MutSigCV) program was implemented to identify genes mutated more frequently than expected by chance, after correcting for patient-specific mutation frequencies and spectra and gene-specific mutation rates, expression levels, and replication times. This analysis identified six significantly mutated genes (Table 2D), consistent with prior studies of ccRCC, including VHL,

PBRM1, and SETD2 (Cancer Genome Atlas Research, 2013). Of these 6, mutations in PBRM1 were more common in extreme responders to anti-PDl therapy than in extreme progressors (p=0.019; Pearson's chi-squared) (Figure 5). It was also noted that some subjects had deletions in various chromosomes. For example, subjects CA209009 12 115 and KE6262 had arm-level monoallelic deletion of chromosome 15, including B2M;

subjects PD_005, PD_007, CA209009_5_1, and CA209009_13_96 had arm-level monoallelic deletion of chromosome 6, including HLA-A, HLA-B, HLA-C, TAP1, TAP2, and TAPBP; subject VA1008 had focal monoallelic deletion of chromosome 6, including HLA-A, HLA-B, HLA-C, TAPl, TAP2, and TAPBP; subject CA209009 8 105 had focal monoallelic deletion of chromosome 6, including HLA-A, HLA-B, and HLA-C; subject CA209009 11 25 had arm-level monoallelic deletion of chromosome 6, including HLA-A, HLA-B, HLA-C, and TAPBP; subject CA209009 11 93 had a large monoallelic deletion of chromosome 6, including HLA-B, HLA-C, and TAPBP; and subject CA209009 5 503 had a large monoallelic deletion of chromosome 6, including HLA-A, HLA-B, TAPl, TAP2, and TAPBP. Furthermore, it was observed that truncating mutations (frameshift indels, nonsense, or splice-site) in PBRMl occurred significantly more frequently in the extreme responders (p= 0.0064; Pearson's chi-squared) after correcting for false discovery among the 6 genes mutated significantly in the training cohort (q=0.039; Benjamini-

Hochberg) (Figure 4C, Table 2E). All truncating PBRMl alterations were in the context of chromosome 3p deletions (Figure 4B), resulting in expected complete loss-of-function of PBRMl . Most of these alterations were predicted to be clonal (present in all tumor cells), with the two subclonal alterations found in one patient with stable disease and another with extreme response to anti-PDl therapy (Table 2E). Patients with truncating mutations in PBRMl had significantly prolonged progression-free survival compared to those without truncating alterations in PBRMl (p = 0.042) (Figure 4D), and prolonged overall survival as well (p = 0.014) (Figure 6), with sustained reductions in tumor burden (Figure 7B). Of note, two of the three extreme progressors with PBRMl truncating mutations had long OS (>1 5 years), and all three were still alive at the time of censoring (Figure 7B).

Additionally, of the three patients with SD and objective tumor regression but PFS of insufficient duration to be considered an exceptional responder, 2 were PBRMl mutants, while the third (2 85) had relatively low tumor sequencing coverage over PBRMl (48-fold) and low tumor purity (estimated 13% tumor cells), making it possible that we were insufficiently powered to detect a PBRMl mutation in this patient. In a focused search for PBRMl alterations in the 6 tumors initially excluded from analysis for quality-control reasons (Figure IB), two additional truncating mutations were found. One was a poorly- supported splice site alteration (4/35 reads, all in reverse direction) in an extreme progressor (4_95), while the other was a well-supported nonsense alteration (22/417 reads) in an extreme responder (5 6).

Table 2C: Summary of Mutational Burden in Training Cohort (N=34)

Table 2D: MutSigCV results in training cohort (N=34)

ATX ataxin 7- 2723 76 1.0 8.25 1.85 2.5 1.17 N7L1 like 1 6E- E-04 E-02 4E- E-05

07 09

GUC guanylat 3326 16 2.0 3.00 4.78 5.6 2.08 Y2C e cyclase 0E- E-03 E-01 8E- E-03 2C (heat 05 07 stable

enteroto

xin

receptor)

KDM lysine 4879 24 4.4 3.50 4.1 1.25 5C (K)- 6E- E-01 0E- E-02 specific 07 06 demethy

lase 5C

On y six identified genes, among 18,345 genes tested are shown in Tab e 2D.

Table 2E: Truncating PBRMl alterations in patients training cohort passing whole exome quality control (N=34)

CA20 P 1 125.64 3 52613 52 Frame T T p.Kl DE 0.4 53 67 1 strel 9009 B 210 61 Shift 146f L 41 ka, 5 21 R 32 Del s 66 indel M 10 66 ocat 1 67 or

CA20 P 1 126.07 3 52678 52 Nonse C c A p.E2 SN 0.1 8 45 1 NA

9009 B 748 67 nse_ 91* P 50

5 18 R 87 Mutat 94

M 48 ion 3

1

CA20 P 1 155.18 3 52620 52 Frame ATTT ATTT p.KI DE 0.0 19 26 0 strel

9009 B 610 62 Shift T T 1087 L 67 4 ka,

5 106 R 06 Del fs 13 indel M 14 78 ocat 1 09 or

CA20 N 0 138.81 N NA NA NA NA NA NA NA NA N NA 9009 A A A 5 1

CA20 N 0 100.73 N NA NA NA NA NA NA NA NA N NA 9009 A A A 4 68

CA20 P 1 94.84 3 52613 52 Nonse c c A p.El SN 0.5 53 47 1 NA

9009 B 194 61 nse_ 105* P 3

3 15 R 31 Mutat

M 94 ion

1

CA20 P 1 146.69 3 52643 52 Nonse G G A p.Q8 SN 0.2 36 89 1 NA 9009 B 375 64 nse_ 09* P 88

3 117 R 33 Mutat

M 75 ion

1

CA20 P 1 111.22 3 52662 52 Frame A A p.N4 DE 0.1 10 82 1 strel 9009 B 964 66 Shift 63fs L 08 ka, 3 114 R 29 Del 69 indel M 64 56 ocat 1 52 or

CA20 N 0 47.52 N NA NA NA NA NA NA NA NA NA NA NA N NA 9009 A A A 2 85

CA20 P 1 130.86 3 52696 52 Frame T T p.Kl DE 0.1 12 58 1 strel

9009 B 272 69 Shift 35fs L 71 ka,

2 84 R 62 Del 42 indel M 72 85 ocat 1 71 or

CA20 N 0 95.31 N NA NA NA NA NA NA NA NA N NA 9009 A A A 2 58

CA20 P 1 266.4 3 52663 52 Splice c c T SN 0.2 25 81 1 NA

9009 B 052 66 _Site P 35

2 102 R 30 84

M 52 9

1

CA20 P 1 164.39 3 52643 52 Frame A A p.S8 DE 0.4 91 13 1 strel

9009 B 489 64 Shift 18fs L 02 5 ka,

13 96 R 34 Del 65 indel M 89 48 ocat 1 67 or

CA20 N 0 124.87 N NA NA NA NA NA NA NA NA N NA 9009 A A A 13 90

CA20 N 0 115.51 N NA NA NA NA NA NA NA NA N NA 9009 A A A 12 11

5

CA20 P 1 173.78 3 52651 52 Splice C C T SN 0.1 6 41 1 NA

9009 B 277 65 _Site P 27

11 93 R 12 66

M 77

1

CA20 P 1 67.19 3 52621 52 Frame T T p.Nl DE 0.4 13 15 1 strel 9009 B 487 62 Shift 017f L 64 ka, 11 79 R 14 Del s 28 indel M 87 57 ocat 1 14 or CA20 N 0 221.56 N NA NA NA NA NA NA NA NA N NA

9009 A A A

11 56

CA20 N 0 124.35 N NA NA NA NA NA NA NA NA N NA

9009 A A A

11 25

CA20 P 1 131.63 3 52623 52 Frame G G - p.D9 DE 0.2 15 45 1 strel

9009 B 201 62 Shift 65fs L 5 ka,

11 14 R 32 Del indel M 01 ocat 1 or

CA20 N 0 62.73 N NA NA NA NA NA NA NA NA N NA

9009 A A A

11 11

CA20 P 1 89.9 3 52623 52 Frame G G - p.I9 DE 0.5 55 45 1 strel

9009 B 120 62 Shift 92fs L 5 ka,

11 10 R 31 Del indel M 20 ocat 1 or

CA20 P 1 131.16 3 52613 52 Splice ACA ACA - DE 0.1 37 17 0 strel

9009 B 062 61 _Site CTC CTC L 73 6 ka

1 62 R 30 A A 70

M 68 89

1 2

CA20 N 0 120.85 N NA NA NA NA NA NA NA NA N NA

9009 A A A

1 32

CA20 P 1 28.98 3 52649 52 Frame - - T p.H6 IN 0.3 8 14 1 strel

9009 B 455 64 JShift 27fs s 63 ka,

1 20 R 94 Ins 63 indel M 56 63 ocat 1 64 or

A validation cohort of 41 patients (see Table 3) treated with immune checkpoint therapy for alterations in PBRMl was then examined to confirm the association between PBRMl mutational status and response to immune checkpoint therapy. After limiting analyses to those treated with immune checkpoint monotherapy and applying the same quality control standards and definitions of clinical response as in the training cohort, PBRMl status was assessed in 28 patients (Figures 7A-7B and Tables 4A-4C). Extreme responders to immune checkpoint therapy were significantly more likely than extreme progressors to harbor truncating alterations in PBRMl (8/13 vs. 1/7, p = 0.043; Pearson's chi-squared) (Figure 7C-7D). Again, all but one truncating event in PBRMl occurred in the setting of chromosome 3p deletion, though this was likely a false negative due to low tumor purity (Figure 8). One patient (VA1008) likely had CN loss over chromose 3p, though low tumor purity made calling this deletion difficult.

In examining germline variants in WES of germline tissue across both the training and validation cohorts (N=91), including samples that failed quality control for tumor WES, 4 nonsynonymous variants (all in extreme responders), but no truncating alterations in PBRMl, were observed (Table 5A). Further analysis covers the frameshift and nonsense variants in genes thought to be associated with hereditary cancer syndromes (Hart et al.

(2016) BMJ Open 6:e010332), as well as genes involved in JAK/STAT signaling and immune checkpoints. Almost all alterations were heterozygous and have been previously observed in a database of germline variants from more than 60 thousand ethnically diverse individuals (ExAC) (Lek et al. (2016) Nature 536:285-291) (Table 5B). Two patients (CA8808: extreme responder and RCC. l lOl : stable disease) had a heterozygous frameshift alteration in PD-L2 (p.LlOfs), which has been observed at frequency of 0.2% in ExAC.

In the somatic space, alterations affecting antigen presentation machinery were rare. In the training cohort, no patients harbored nonsynonymous alterations in TAP1, TAP2, B2M, TAPBP, or any of the HLA Class I alleles. One patient with intermediate benefit (12 115) had a heterozygous deletion of B2M. Six patients (2 extreme responders, 1 intermediate benefit, 2 extreme progressors) had loss of heterozygosity (LOH) in chromosome 6p affecting the HLA and TAP loci. In the validation cohort, one patient had a nonsense mutation in TAPl (VA1008; extreme responder), one had a missense mutation in B2M (PD 021; extreme progressor), and one had LOH of B2M (KE6262; extreme responder). Three patients had LOH over chromosome 6p (2 extreme responders, 1 intermediate benefit; see also Figure 12).

Table 3 Clinical cohort consolidation

Table 4A: Sequencing Metrics and Inclusion/Exclusion Criteria for Whole Exome

Sequencing in Validation Cohort (N=41)

RCC- 135.674171 100.347012 whole exome illumina c 0.4 3.78 1 CombinationT PD 004 oding vl KI

RCC- 149.115421 72.503658 whole exome illumina c 0.5 2.37 0 CombinationT

PD 006 oding vl KI

RCC- 159.873929 91.682176 whole exome illumina c 0.39 1.95 0 CombinationT PD 008 oding vl KI

RCC- 180.675064 91.140713 whole exome illumina c 0.45 2.99 1 CombinationT

PD 027 oding vl KI

RCC- 150.285278 98.786695 whole exome illumina c 0.25 1.79 0 CombinationT PD 028 oding vl KI

RCC- 178.994864 93.923124 whole exome illumina c 0.55 2.07 0 CombinationT PD 031 oding vl KI

RCC- 106.860416 72.680181 whole exome illumina c 0.5 1.89 0 CombinationT

PD 002 oding vl KI

RCC- 125.608438 102.090575 whole exome illumina c 0.47 1.82 0 0 PD 005 oding vl

RCC- 141.661729 101.362659 whole exome illumina c 0.57 1.82 0 0

PD 007 oding vl

RCC- 122.598167 87.563055 whole exome illumina c 0.41 1.99 0 PapillaryRCC

PD 009 oding vl

RCC- 104.135516 82.878525 whole exome illumina c 0.32 2.08 0 0 PD Oil oding vl

RCC- 133.950619 87.549415 whole exome illumina c 0.39 2.01 0 0 PD 012 oding vl

RCC- 145.082205 91.170952 whole exome illumina c 0.22 3.44 1 0 PD 013 oding vl

RCC- 127.306107 78.539083 whole exome illumina c 0.25 4.02 1 0 PD 014 oding vl

RCC- 105.708638 93.290512 whole exome illumina c 0.36 2.1 0 0 PD 015 oding vl

RCC- 145.443729 95.372761 whole exome illumina c 0.61 1.84 0 0 PD 018 oding vl

RCC- 148.823821 87.774525 whole exome illumina c 0.42 1.97 0 0 PD 019 oding vl

RCC- 151.788377 102.972091 whole exome illumina c 0.18 1.86 0 0

PD 020 oding vl

RCC- 159.181781 95.98438 whole exome illumina c 0.58 1.9 0 0 PD 021 oding vl

RCC- 148.651377 98.276519 whole exome illumina c 0.53 1.97 0 0

PD 022 oding vl

RCC- 135.431357 86.807511 whole exome illumina c 0.19 2.65 1 0

PD 023 oding vl

RCC- 92.006306 83.700183 whole exome illumina c 0.27 2.01 0 0

PD 024 oding vl

RCC- 74.164294 48.102291 whole exome illumina c 0.35 1.99 0 0 PD 025 oding vl

RCC- 166.502187 89.436443 whole exome illumina c 0.76 1.98 0 0

PD 026 oding vl

CA8808 Tl 123.07315 103.475727 whole exome agilent 1 0.43 1.96 0 0

KA4076 Tl 126.229037 120.209259 whole exome agilent 1 0.56 2.03 0 0

KE5236 Tl 132.886302 140.196056 whole exome agilent 1 0.33 2 0 0

KE6262 Tl 99.539361 106.858872 whole exome agilent 1 0.11 4.16 1 0

MC1838 Tl 149.730846 118.307339 whole exome agilent 1 0.41 1.99 0 0

VA1008 Tl 142.542157 89.429498 whole exome agilent 1 0.14 1.74 0 0

RCC.PD1.D 92.348009 81.023695 whole exome agilent l. 0.3 1.97 0 0 NA.1101.T 1 refseq plus 3 boosters

RCC.PD1.D 71.474257 96.238769 whole exome agilent l. 0.31 3.43 1 0 NA.1137.T 1 refseq_plus 3 boosters

RCC.PD1.D 136.955167 87.050978 whole exome agilent l. 0.26 2.15 0 0 NA.1026.T 1 refseq_plus 3 boosters

RCC.PD1.D 126.472115 95.64198 whole exome agilent l. 0.66 1.89 0 0 NA.944.T 1 refseq plus 3 boosters

RCC.PD1.D 101.276419 96.351667 whole exome agilent l. 0.43 3.91 1 0 NA.949.T 1 refseq plus 3 boosters

Table 4B: Clinical Information for Immune-Checkpoint- Treated Patients in Validation

Cohort (N=41) eci cha nso une

St nge r che

ckp

oint

VA1008 nivolumab + PR M 76 -96 clear- 1135 1 1135 1 0 extreme re extreme re ipilimumab cell sponder sponder

RCC.PD nivolumab PR F 60 -40 clear- 364 0 235 0 0 extreme re extreme re 1.DNA.9 cell sponder sponder 49

RCC.PD nivolumab PD M 47 37 clear- 134 1 67 0 0 extreme_pr extreme_pr 1.DNA.9 cell ogressor ogressor 44

RCC.PD nivolumab SD F 61 16 clear- 1584 1 119 0 0 stable dise stable dise 1.DNA.1 cell ase ase 137

RCC.PD nivolumab SD M 67 4 clear- 439 0 171 0 0 stable dise stable dise 1.DNA.1 cell ase ase 101

RCC.PD nivolumab CR M 60 -87 clear- 1442 1 357 1 0 extreme re extreme re 1.DNA.1 cell sponder sponder 026

RCC- axitinib + PR M 68 -49 clear- 165 1 123 0 0 extreme re stable dise PD 031 avelumab cell sponder ase

RCC- nivolumab X M 72 -43 clear- 395 0 93 0 0 not evalua stable dise

PD 030 cell ble ase

RCC- nivolumab PR M 54 -49 clear- 856 0 189 0 0 extreme re extreme re

PD 029 cell sponder sponder

RCC- atezolizumab PR M 77 -43 clear- 210 1 210 1 0 extreme re extreme re PD 028 + cell sponder sponder bevacizumab

RCC- axitinib + PR M 59 -42 clear- 210 1 210 1 0 extreme re extreme re

PD 027 avelumab cell sponder sponder

RCC- nivolumab SD F 70 20 clear- 377 1 171 0 0 stable dise stable dise

PD 026 cell ase ase

RCC- nivolumab SD M 74 -23 clear- 1724 1 333 0 0 stable dise extreme re PD 025 cell ase sponder

RCC- nivolumab PD M 52 30 clear- 304 0 41 0 0 extreme_pr extreme_pr

PD 024 cell ogressor ogressor

RCC- atezolizumab PR M 69 -88 clear- 637 1 637 1 0 extreme re extreme re

PD 023 cell sponder sponder

RCC- nivolumab PD F 66 NA clear- 247 1 80 0 0 extreme_pr extreme_pr

PD 022 cell ogressor ogressor

RCC- nivolumab PD F 63 NA clear- 185 0 68 0 0 extreme_pr extreme_pr PD 021 cell ogressor ogressor

RCC- nivolumab PD F 64 NA clear- 203 1 47 0 0 extreme_pr extreme_pr

PD 020 cell ogressor ogressor

RCC- nivolumab SD M 60 -11 clear- 230 1 220 0 0 stable dise extreme re PD 019 cell ase sponder

RCC- nivolumab PR F 69 -82 clear- 1189 0 672 0 0 extreme re extreme re PD 018 cell sponder sponder

RCC- nivolumab PD M 71 6 clear- 814 0 105 0 0 stable dise stable dise PD 015 cell ase ase

RCC- nivolumab + SD F 68 -5 clear- 433 1 433 1 0 extreme re extreme re PD 014 ipilimumab cell sponder sponder

RCC- nivolumab + PR M 66 -32 clear- 399 1 399 1 0 extreme re extreme re PD 013 ipilimumab cell sponder sponder

RCC- atezolizumab PD M 67 -50 clear- 581 1 61 0 0 extreme_pr extreme_pr PD 012 cell ogressor ogressor

RCC- nivolumab PD M 40 -37 clear- 327 0 205 0 0 stable dise extreme re PD Oil cell ase sponder

RCC- nivolumab + CR M 51 -51 clear- 454 1 454 1 0 extreme re extreme re PD 010 ipilimumab cell sponder sponder

RCC- nivolumab + PD M 56 8 papill 377 1 89 0 0 extreme_pr extreme_pr

PD 009 ipilimumab ary ogressor ogressor

RCC- axitinib + PR F 69 -69 clear- 462 1 462 1 0 extreme re extreme re PD 008 pembrolizuma cell sponder sponder b

RCC- nivolumab + PR M 60 -42 clear- 448 1 448 1 0 extreme re extreme re

PD 007 ipilimumab cell sponder sponder

RCC- axitinib + PR M 68 -52 clear- 398 1 398 1 0 extreme re extreme re

PD 006 pembrolizuma cell sponder sponder b RCC- nivolumab PD M 62 NA clear- 277 1 168 0 0 stable dise stable dise PD 005 cell ase ase

RCC- axitinib + SD M 54 -16 clear- 481 1 481 1 0 extreme re extreme re PD 004 pembrolizuma cell sponder sponder b

RCC- atezolizumab SD M 52 -16 clear- 679 1 479 0 0 extreme re extreme re

PD 003 + cell sponder sponder bevacizumab

RCC- atezolizumab SD M 65 -14 clear- 534 1 255 0 1 stable dise extreme re

PD 002 + cell ase sponder bevacizumab

RCC- axitinib + PR F 66 -53 clear- 572 1 572 1 0 extreme re extreme re PD 001 pembrolizuma cell sponder sponder b

MC1838 nivolumab PD M 64 93 clear- 622 0 60 0 0 extreme_pr extreme_pr cell ogressor ogressor

KE6262 nivolumab PR M 68 -60 clear- 903 1 163 0 0 extreme re stable dise cell sponder ase

KE5236 nivolumab PD M 58 70 clear- 997 1 165 0 0 stable dise stable dise cell ase ase

KA4076 nivolumab PD F 61 59 clear- 727 0 107 0 0 stable dise stable dise cell ase ase

CA8808 nivolumab PR M 62 -55 clear- 560 1 558 0 0 extreme re extreme re cell sponder sponder

BL5166 nivolumab SD M 64 -11 clear- 622 0 156 0 0 stable dise stable dise cell ase ase

For sex, M represents male and F represents female.

Table 4C: Truncating PBRMl alterations in validation cohort

PD 01 PBR 3 5263 5263 Frame AG AG p.R9 DEL 0.20 18 70 1 strelka, 5 Ml 7540 7540 Shift 41fs 4545 indeloc

Del 455 ator

PD 01 NA NA NA NA NA NA NA NA NA NA NA NA 8

PD 01 NA NA NA NA NA NA NA NA NA NA NA NA

9

PD 02 NA NA NA NA NA NA NA NA NA NA NA NA 0

PD 02 PBR 3 5271 5271 Frame C C p.G2 DEL 0.36 18 32 1 strelka, 1 Ml 3723 3723 Shift fs indeloc

Del ator

PD 02 NA NA NA NA NA NA NA NA NA NA NA NA

2

PD 02 PBR 3 5266 5266 Splice T T A SNP 0.21 9 33 1 NA

3 Ml 3053 3053 Site 4286

PD 02 NA NA NA NA NA NA NA NA NA NA NA NA 4

PD 02 PBR 3 5259 5259 Frame c c p.Gl DEL 0.15 21 115 1 strelka, 5 Ml 5829 5829 Shift 429f 4411 indeloc

Del s 765 ator

PD 02 NA NA NA NA NA NA NA NA NA NA NA NA 6

RCC.P NA NA NA NA NA NA NA NA NA NA NA NA Dl.DN

A.1026

RCC.P PBR 3 5259 5259 Frame c c p.Al DEL 0.13 43 279 1 strelka, Dl.DN Ml 5804 5804 Shift 438f 3540 indeloc A.1101 Del s 373 ator

RCC.P NA NA NA NA NA NA NA NA NA NA NA NA Dl.DN

A.1137

RCC.P NA NA NA NA NA NA NA NA NA NA NA NA Dl.DN

A.944

RCC.P NA NA NA NA NA NA NA NA NA NA NA NA Dl.DN

A.949

VA100 PBR 3 5264 5264 Frame T T p.K6 DEL 0.06 15 253 not indeloc 8 Ml 3943 3943 Shift 19fs evalu ator

Del able

Table 4D: Truncating PBRMl alterations in patients receiving immune checkpoint therapy in combination with angiogenesis inhibitor or tyrosine kinase inhibitor in validation cohort (N=9)

PD NA NA NA NA NA NA NA NA NA NA NA NA NA NA

002

Table 5 A: Germline variants in PBRMl in training and validation cohorts (N=91)

All samples had germline variations (SNPs) on PBRMI (Chrom. 3).

Table 5B: Germline variants in cancer susceptibility genes in training and validation cohorts (N=91)

CA8 PDC 9 5522 55225 G G - Frame DE c.30delG c.(28- p.Ll 1 1 0.0

808 D1LG 576 76 Shift L 30)ctgf Ofs 1 0 01

2 Del s 0 7 98

5

CA2 FAN 2 5838 58386 - - TA Frame IN c.1114 1 c.(1114 p.T3 3 5 0.0

0900 CL 6928 929 AT JShift S 115insA - 72fs 8 3 02

9 5 Ins TTA 1116)a 83

22 ccfs 4

PD BRC 1 3297 32972 A A T Nonse SN .9976A c.(9976 p.K3 4 4 0.0

Oil A2 3 2626 626 nse_ P >T - 326* 9 8 07

Mutat 9978)A 01 ion aa>Taa

PD BRC 1 3297 32972 A A T Nonse SN .9976A c.(9976 p.K3 6 4 0.0

003 A2 3 2626 626 nse_ P >T - 326* 1 2 07

Mutat 9978)A 01 ion aa>Taa

PD MSR 8 1601 16012 G G A Nonse SN c.931C> c.(931- p.R3 4 5 0.0

013 1 2594 594 nse_ P T 933)Cg 11* 5 0 07

Mutat a>Tga 34 ion 8

CA2 AR X 6676 66766 GG GG - In Fr DE 1 2 0.0

0900 6357 374 CG CG ame L 0 3 14

9 9 GC GC Del 4 3 89

52 GG GG

CG CG

GC GC

RCC GEN1 2 1796 17962 AA AA - Frame DE c.2515 2 c.(2515 p.K rsl49 2 3 0.0

.PD 2994 998 GT GTT Shift L 519delA - L83 93694 7 1 88

1.D T Del AGTT 2520)a 9fs 4

NA. agttgfs

1101

RCC GEN1 2 1796 17962 AA AA - Frame DE c.2515 2 c.(2515 p.K rsl49 2 3 0.0

.PD 2994 998 GT GTT Shift L 519delA - L83 93694 2 4 88

1.D T Del AGTT 2520)a 9fs 4

NA. agttgfs

1082

RCC GEN1 2 1796 17962 AA AA - Frame DE c.2515 2 c.(2515 p.K rsl49 5 0 0.0

.PD 2994 998 GT GTT Shift L 519delA - L83 93694 9 88

1.D T Del AGTT 2520)a 9fs 4

NA. agttgfs

1026

PD GEN1 2 1796 17962 AA AA - Frame DE c.2515 2 c.(2515 p.K rsl49 3 3 0.0

027 2994 998 GT GTT Shift L 519delA - L83 93694 2 2 88

T Del AGTT 2520)a 9fs 4

agttgfs

PD GEN1 2 1796 17962 AA AA - Frame DE c.2515 2 c.(2515 p.K rsl49 4 3 0.0

026 2994 998 GT GTT Shift L 519delA - L83 93694 1 6 88

T Del AGTT 2520)a 9fs 4

agttgfs

PD GEN1 2 1796 17962 AA AA - Frame DE c.2515 2 c.(2515 p.K rsl49 3 2 0.0

009 2994 998 GT GTT Shift L 519delA - L83 93694 0 7 88

T Del AGTT 2520)a 9fs 4

agttgfs

PD GEN1 2 1796 17962 AA AA - Frame DE c.2515 2 c.(2515 p.K rsl49 2 3 0.0

006 2994 998 GT GTT Shift L 519delA - L83 93694 2 6 88

T Del AGTT 2520)a 9fs 4

agttgfs

PD GEN1 2 1796 17962 AA AA - Frame DE c.2515 2 c.(2515 p.K rsl49 3 3 0.0

004 2994 998 GT GTT Shift L 519delA - L83 93694 8 5 88

T Del AGTT 2520)a 9fs 4

agttgfs

PD GEN1 2 1796 17962 AA AA - Frame DE c.2515 2 c.(2515 p.K rsl49 5 0 0.0

002 2994 998 GT GTT Shift L 519delA - L83 93694 0 88

T Del AGTT 2520)a 9fs 4

agttgfs

KA4 GEN1 2 1796 17962 AA AA - Frame DE c.2515 2 c.(2515 p.K rsl49 5 5 0.0

0761 2994 998 GT GTT Shift L 519delA - L83 93694 7 3 88

T Del AGTT 2520)a 9fs 4

agttgfs

CA2 GEN1 2 1796 17962 AA AA - Frame DE c.2515 2 c.(2515 p.K rsl49 4 0 0.0

0900 2994 998 GT GTT Shift L 519delA - L83 93694 7 88

9 9 T Del AGTT 2520)a 9fs 4

45 agttgfs

To further characterize the effect of PBRMl truncating alterations on the tumor- immune microenvironment, publicly available genomic data from the Cancer Genome Atlas (TCGA) clear-cell RCC (KIRC) cohort with matched whole exome and whole transcriptome sequencing (Cancer Genome Atlas Research, 2013) were analyzed. A study of immune checkpoint expression in patient samples as well as in TCGA KIRC showed that the expression profiles of multiple cytokines were not significantly changed by PBRMl truncation (Figure 9).

Table 5C

s nga = nt

This study found that patients with truncating alterations in PBRMl are more likely to experience extreme response to immune checkpoint monotherapy than patients who are PBRMl -wildtype (Figure 10). Meanwhile, nonsynonymous mutational burden, neoantigen burden, and PD-L1 staining did not distinguish clinical benefit groups, in contrast to findings in melanoma and non-small cell lung cancer (Rizvi et al. (2015) Science 348: 124- 128; Snyder et al. (2014) N. Engl. J. Med. 371 :2189-2199; Van Allen et al. (2015) Science 350:207-211).

PBRMl is a component of the BAF (Brg/Brahma-associated factors) or mammalian SWI/S F complex, which is involved in ATP-dependent chromatin remodeling, and is one of the most commonly mutated genes in ccRCC. Nonsynonymous mutations in PBRMl are seen in up to 41% of patients with ccRCC (Varela et al. (2011) Nature 469:539-542), with a majority of mutations being truncating alterations. Chromosome 3p deletions over the PBRMl locus are also highly prevalent in ccRCC (>91% of samples), as are alterations in other components of the SWI/SNF complex, including BAPl and SETD2, suggesting that epigenetic regulation and oncogenic metabolism are major components of ccRCC (Cancer Genome Atlas Research (2013), supra). The tumor suppressor role of PBRMl loss in ccRCC is most often associated with metabolism, hypoxia response, and cell adhesion (Chowdhury et al. (2016), supra) but it may have interesting effects on the tumor-immune microenvironment as well.

Additionally, restoration of PBRMl expression in PBRMl -deficient tumor cell lines leads to increased expression of genes in the interleukin-6-mediated signaling pathway (GO:0070102) (Chowdhury et al. (2016), supra, and was observed lower levels of IL-6 in the baseline serum of patients with PBRMl -truncated tumors in this study. Increased production of IL-6 mediates STAT3 activation, which has been identified as a potential orchestrator of an immunosuppressive cytokine network (Yu et al. (2009) Nat. Rev. Cancer 9:798-809), and promotes tumorigenesis in EGFR-mutant lung carcinomas (Gao et al. (2007) J. Clin. Invest. 117:3846-3856). Activation of the Jak2/Stat3 pathway has been further associated with an immunosuppressive tumor microenvironment in Pten-mxW mice that develop prostatic neoplasia, and blockade of this pathway can restore the anti-tumor immune response (Toso et al. (2014) Cell Rep. 9:75-89). This finding is further supported by decreased macrophage and T cell infiltration in PBRM1 -truncated tumors, along with decreased CRP and increased IP- 10 immediately before treatment.

Taken together, these results indicate that PBRM1 status may have wide-ranging effects on tumor-immune microenvironment interactions. Clinically, alterations in PBRM1 have previously been linked with prognosis and response to other cancer therapies. A possibility cannot be fully excluded that PBRM1 has prognostic rather than predictive value. One study in 145 patients found that PBRM1 -mutant tumors were associated with favorable prognosis, especially relative to BAP 1 -mutant tumors (Kapur et al. (2013), supra), while another study in 609 patients found no effect of PBRM1 mutations on cancer- specific survival (Hakimi et al. (2013) Clin. Cancer Res. 19:3259-3267). These studies did not distinguish between truncating and non-truncating (missense mutations, in-frame indel) variants or assess chromosome 3p.21 deletions, which could impact the ultimate presence of PBRM1 protein. Immunohistochemical staining for PBRM1 in 657 ccRCC cases found worse cancer-specific survival and progression-free survival in poorly staining samples (Nam et al. (2015) Urol. Oncol. 33 :340. e9-el6), and a similar study in 204 ccRCC cases also found that loss of PBRM1 protein expression is associated with poor differentiation, late tumor stage, and shorter duration of patient overall survival (Pawlowski et al. (2013) Int. J. Cancer 132:E11-E17).

Previous studies have also investigated whether pre-treatment molecular

characteristics of ccRCC are correlated to response to therapy. In a cohort of 258 patients with RCC, those with PBRMl -mutant cancers were found to have longer PFS with first- line everolimus compared to those who were PBRMl -wildtype, though this finding did not hold after multiple hypothesis testing (Hsieh et al. (2016) Eur Urol, pii: S0302-

2838(16)30701-1). No effect of PBRMl status was seen with first-line sunitinib followed by everolimus in the same trial. Another study in 27 patients treated with vascular endothelial growth factor (VEGF) targeted therapies (sunitinib and pazopanib) found that PBRMl alterations were significantly enriched in responders (Fay et al. (2016) J. Natl. Compr. Cane. Netw. 14:820-824), while a third study in cohort of 79 patients receiving mTOR inhibitors (everolimus and temsirolimus) found no association between PBRMl status and response (Kwiatkowski et al. (2016) Clin. Cancer Res. 22:2445-2452). Another study including 117 pre-treatment tumors found no association between somatic mutations in PBRMl and response to sunitinib (Beuselinck et al. (2015) Clin. Cancer Res. 21 : 1329- 1339). Thus, the observed association between PBRMl mutations and increased likelihood of clinical benefit from immune checkpoint therapy is a novel finding not readily explained by general decreased tumor aggressiveness or increased responsiveness to therapy in PBRMl -mutant tumors. Additionally, all extreme responders in this study were required to have objective decrease in tumor burden following immune checkpoint therapy, making it unlikely that the prognostic benefit of PBRMl mutation alone, if real, could explain the results of this study.

This finding of increased responsiveness to immune checkpoint therapy in patients with metastatic ccRCC harboring truncating mutations in PBRMl in independent training and validation cohorts totaling 61 patients argues for further validation in larger

immunotherapy -treated RCC cohorts and for concerted effort towards characterizing the impact of SWI/S F complex alterations on tumor-immune activity. Integration of whole exome and whole transcriptome sequencing from patient tumors identified potential downstream effects of PBRMl alterations on immune cell infiltration. These results are believed to have important implications for exploration of PBRMl and immune mediation, as well as guiding patient selection for immune checkpoint therapy in renal cell carcinoma, where up to 40% of patients have PBRMl-mutant disease. This finding may is also believed tobe more generally relevant in cancer immunotherapy, as more than 20% of human cancers contain a mutation in at least one subunit of the SWI/SNF or BAF complexes (Kadoch et al. (2013) Nat. Genet. 45:592-601; Shain and Pollack (2013) PLoS One 8:e55119). These results can be further applied to untreated RCC cohorts (Sato et al. (2013), supra).

Example 3: Further Confirmation of Data and Results Shown In Examples 1-2

The following provide further confirmation of the data and results provided above in Examples 1-2 by inter alia further demonstrating the data and results in additional cohorts. Generally, the following materials and methods were used to determine the further confirmation:

a. Clinical cohort consolidation

The discovery cohort was gathered from patients enrolled in p-009 (NCTO 1358721), a study of nivolumab (BMS-936558) monotherapy in metastatic renal cell carcinoma (Choueiri et al. (2016) Clin. Cancer Res. 22:5461-5471). Progression-free survival and overall survival were measured from Cycle 1 Day 1 (time zero) of nivolumab

administration. The validation cohort was gathered from patients at the Dana-Farber Cancer Institute, Memorial Sloan Kettering Cancer Institute, and Johns Hopkins University who received anti-PD-(L)l therapy as monotherapy or in combination with other immune checkpoint therapies and had banked adequate pre-treatment tumor tissue for molecular characterization. In addition, patients with ccRCC also treated with anti-PD-(L)l based therapy from the Mayo Clinic with targeted panel sequencing that included the PBRM1 gene region were included in the validation cohort. All patients were consented on an Institutional Review Board protocol that allows research molecular characterization of tumor and germline samples. Each IRB at the respective institution from the validation cohort obtained approval for 1) collection and analysis of samples, and 2) sending samples to the Dana-Farber Center for genomic analysis.

b. DNA and RNA extraction and sequencing

All samples from the discovery cohort and those from the Dana-Farber Cancer

Institute and Memorial Sloan Kettering Cancer Institute were processed for DNA (and if possible, RNA) extraction and whole exome sequencing through standard workflows (Van Allen et al. (2014) Nat. Med. 20:682-688 ). After fixation and mounting, 5-10 10 μιη slices from either Qiagen RNAlater (discovery cohort) or formalin-fixed, paraffin-embedded (FFPE, validation cohort) tumor blocks were obtained, and tumor-enriched tissue was macrodissected. Paraffin was removed from FFPE sections and cores using CitriSolv™ (Fisher Scientific), followed by ethanol washes and tissue lysis overnight at 56°C. Samples were then incubated at 90°C to remove DNA crosslinks, and DNA- and when possible, RNA-extraction was performed using Qiagen AllPrep DNA/RNA Mini Kit (#51306).

Germline DNA was obtained from adjacent PBMCs. Whole exome and whole

transcriptome sequencing of tumor and germline samples were performed as previously described in Van Allen et al. (2015) Science 350:207-211 and Van Allen et al. (2014) Nat. Med. 20:682-688. All samples in the discovery cohort were sequenced using the Illumina exome, while a portion of the samples in the validation cohort were sequenced using the Agilent exome (Table 6E). The Illumina exome uses Illumina' s in-solution DNA probe based hybrid selection method to target approximately 37.7Mb of mainly exonic territory, using similar principles as the Broad Institute- Agilent Technologies developed in- solution RNA probe based hybrid selection method (Agilent SureSelect All Exon V2)

(Gnirke et al. (2009) Nat. Biotechnol. 27: 182-189; Fisher et al. (2011) Genome Biol. 12:R1) to generate Illumina exome sequencing libraries. Pooled libraries were normalized to 2 nM and denatured using 0.2 N NaOH prior to sequencing. Flowcell cluster amplification and sequencing were performed according to the manufacturer's protocols using either the HiSeq 2000 v3 or HiSeq 2500. Each run was a 76 bp paired-end with a dual eight-base index barcode read. Data were analyzed using the Broad Picard Pipeline, which includes de-multiplexing and data aggregation. Exome sequence data processing was performed using established analytical pipelines at the Broad Institute. A BAM file was produced using the Picard pipeline (available on the World Wide Web at picard.sourceforge.net/), which aligns the tumor and normal sequences to the hgl9 human genome build using

Illumina sequencing reads. The BAM was uploaded into the Firehose pipeline (available on the World Wide Web at broadinstitute.org/cancer/cga/Firehose), which manages input and output files to be executed by GenePattern (Reich et al. (2006) Nat. Genet. 38:500- 501). Samples with mean target coverage less than 25x in the tumor and less than 15x in matched normal were excluded. Quality control modules within Firehose were applied to all sequencing data for comparison of the origin of tumor and normal genotypes and to assess fingerprinting concordance. Cross-contamination of samples was estimated using ContEst (Cibulskis et al. (2011) Bioinform. 27:2601-2602). Samples with ContEst estimates exceeding 5% were excluded from analysis. Clinical characteristics from samples that were excluded due to poor quality did not differ significantly from those that were included in the final analysis.

c. Whole exome and whole transcriptome analyses

MuTect was applied to identify somatic single-nucleotide variants (Cibulskis et al. (2013) Nat. Biotechnol. 31 :213-219). Strelka was used to identify somatic insertions and deletions (Saunders et al. (2012) Bioinform. 28: 1811-1817) across the whole exome.

Indelocator, which detects small insertions and deletions after local realignment of tumor and normal sequences, was additionally applied to provide further sensitivity to detect indels in PBRMl (Cancer Genome Atlas Research (2011) Nature 474:609-615). The union of indels called by Strelka and Indelocator was used for final analysis. Artifacts introduced by DNA oxidation during sequencing were computationally removed using a filter-based method (Costello et al. (2013) Nuc. Acids Res. 41 :e67). All somatic mutations detected by whole-exome sequencing were analyzed for potential false positive calls by performing a comparison to mutation calls from a panel of 2,500 germline DNA samples (Stachler et al. (2015) Nat. Genet. 47: 1047-1055). Mutations found in germline samples were removed from analysis. Annotation of identified variants was done using Oncotator (available on the World Wide Web at broadinstitute.org/cancer/cga/oncotator). All nonsynonymous alterations in PBRM1 were manually reviewed in Integrated Genomics Viewer

(IGV 2.3.57) for sequencing quality (Thorvaldsdottir et al. (2013) Brief Bioinform. 14: 178- 192). PBRM1 LOF events were defined as truncating mutations: nonsense mutations, frameshift insertions and deletions, and splice-site mutations. In-frame insertions and deletions, missense mutations, and other alterations presumed not to be truncating were considered separately. Copy ratios were calculated for each captured target by dividing the tumor coverage by the median coverage obtained in a set of reference normal samples. The resulting copy ratios were segmented using the circular binary segmentation algorithm (Olshen et al. (2004) Biostatistics 5:557-572). Allelic copy number alterations were called while taking into account sample-specific overall chromosomal aberrations (focality) (Brastianos et al. (2015) Cancer Discov. 5: 1164-1177). Inference of mutational clonality, tumor purity, and tumor ploidy was accomplished with ABSOLUTE (Carter et al. (2012) Nat. Biotechnol. 30:413-421). Mutations were considered clonal if the expected cancer cell fraction (CCF) of the mutation as estimated by ABSOLUTE was 1, or if the probability of the mutation being clonal was greater than that of the mutation being subclonal. For the discovery cohort, samples were required to have estimated tumor purity greater than 10% to be included in the final analysis. For the validation cohort, samples included in the analysis were required to have either (a) estimated tumor purity greater than 10%, or (b) estimated tumor purity below 10% but sufficient sequencing coverage over the PBRM1 region that there would still be adequate power to detect a clonal PBRM1 alteration if it were to exist. As a final quality control metric to ensure adequate sequencing coverage and tumor purity to detect relevant oncogenic events, all samples had to have at least one nonsynonymous mutation in at least one high confidence or candidate cancer driver gene to be included in the final analysis (Tamborero et al. (2013) Sci. Rep. 3 :2650). Mutation calls for patients from patients from Johns Hopkins University included in the validation cohort were processed through in-house standard analytic pipelines and supplied by Mark Ball, MD (Anagnostou et al. (2017) Cancer Disc. 7:264-276).

d. Targeted sequencing analyses

Fourteen samples with targeted panel genetic sequencing were used in the validation cohort. Panel sequencing data was acquired using standard pipelines from commercial molecular profiling laboratories: FoundationOne® (Foundation Medicine, Palo Alto, CA) and Caris Molecular Intelligence (Cans Life Sciences, Phoenix, AZ) (Table 6E). A subset of these samples had PBRM1 immunohistochemical staining (IHC, Table 6G. All samples with canonical LOF mutations (frameshift insertions, frameshift deletions, splice site mutations) and available PBRM1 IHC had negative staining, indicating true PBRM1 LOF. One patient (MCA6) with missense mutation N258S, also had negative IHC staining, and was labeled a PBRMl-LO¥ mutant accordingly.

e. Cell line analysis

Whole transcriptome sequencing from PBAF-deficient and PBAF-proficient A704 cell lines was produced as previously described in Gao et ol. (2017) Proc. Natl. Acad. Sci. USA 114: 1027-1032 and is available on Gene Expression Omnibus (GEO) under Accession PRJNA371283. Differential gene expression analysis was conducted using the

Bioconductor software package Empirical Analysis of Digital Gene Expression Data in R (edgeR). This package is optimized for differential expression analysis of RNA-seq data with biological replication. Raw read count data from RNA-seq analysis of two PBRMl - null cell lines, two BRGl-null cell lines, and two PBRMl- and BRGl-wild type cell lines were analyzed for differential expression between PBRMl -null (A704) and wildtype (A704BAF180wt), and BRGl-null (A704BAF180wt, BRG1-/-) and wild type cell lines (A704BAF180wt). In order to assess PBAF complex functionality as a whole, the top 100 positively differentially expressed genes by quasi-likelihood F test in mutants vs. wild type from both PBRMl and BRG1 analyses were intersected to get a final list of 48 genes significantly up-regulated in PBAF null cell lines. The same analysis was performed for the top 100 negatively differentially expressed genes, and the resulting list was 43 genes significantly up-regulated in PBAF wild type cell lines. GSEA (available on the World Wide Web at software.broadinstitute.org/gsea/index.jsp) was performed to test whether any biologically-relevant gene sets were differentially expressed between PBAF-null vs.

wildtype and BRG1 null vs. wildtype cell lines. In accordance with previously proposed methods in Liberzon et al. (2015) Cell Sys. 1 :417-425, the Hallmark gene sets (N = 50) were used for an initial GSEA run, and subsequent GSEA analyses were conducted using the Founders gene sets for any Hallmark gene set significantly enriched in both PBRM1 and BRG1 null cell lines (N = 5). A false discovery rate (FDR) q-value of 0.25 was used as a significance threshold for all analyses. This process resulted in a list of gene sets significantly enriched in PBAF-null vs. wildtype cell lines. GSEA analyses were repeated for RNA-Seq from untreated patient tumors from the TCGA. Gene Ontology (GO, available on the World Wide Web at geneontology.org/) term analysis was performed to identify pathways or functional associations of the core enriched genes in A704BAF180-/- versus A704BAF180wt from the Kegg Cytokine-Cytokine Receptor Interaction gene set. Core enriched genes for A704BAF180-/- (N = 53) were defined as those with a GSEA enrichment score greater than the prior gene, starting from the top of the GSEA ranked gene list (i.e., all genes until the peak of the GSEA enrichment plot). Core enriched genes for A704BAF180wt (N = 18) were those whose enrichment score was less than the prior gene, starting from the bottom of the GSEA ranked list (i.e., all genes after the trough of the GSEA enrichment plot).

f . Transcriptome analysis

Whole transcriptome sequencing was derived from three sources: patient samples from the discovery and validation cohorts, the TCGA clear cell renal cell carcinoma (KTRC) cohort, the TCGA cutaneous melanoma (SKCM) cohort, and an independent previously published cohort of untreated clear cell renal cell carcinoma tumors (Sato) (Sato et al. (2013) Nat. Genet. 45:860-867). For the patient samples, whole transcriptome sequencing from FFPE tissues were aligned using STAR (Dobin et al. (2012) Bioinform. 29: 15-21) and then quantified with RSEM (Li et al. (2011) BMC Bioinform. 12:323) to yield gene-level expression in transcripts per million (TPM). Because patient samples came from two independent cohorts, ComBat (Li et al. (2011) BMC Bioinform. 12:323) was applied prior to analyzing patient-derived RNA sequencing. Principal components analysis (PCA) was completed before and after implementing ComBat to ensure that batch effects were eliminated (Johnson et al. (2007) Biostat. 8: 118-127). The final patient cohort for RNA-seq analysis included N= 18 PBRM1-LOF samples and N=14 PBRM1 -intact samples. For the TCGA cohort, whole exome mutation annotation files (MAFs) and whole transcriptome gene expression data were downloaded from the Firebrowse KTRC TCGA data release (2016 01 28). KIRC tumors were divided into those with truncating mutations m PBRMl (nonsense, splice-site, frameshift) (N=102), those with intact PBRM1 function (no mutation or silent mutation) (N=288), and those with other mutations in PBRM1 (missense or inframe indel) (N=25). RNA-seq from germline samples was excluded. For the Sato cohort, whole exome mutation annotation files and gene expression data from the final analysis in the published paper were used (Sato et al. (2013) Nat. Genet. 45:860-867). The MAFs were downloaded from the online supplemental materials from the published paper and gene expression data were kindly supplied by personal

communication with the authors,

g. Statistical analyses

All comparisons of continuous variables between groups (clinical benefit vs. no clinical benefit or PBRM1 -LOF vs. PBRM1 -intact) were done with the non-parametric Wilcoxon rank-sum test (wilcox.testQ R function, two-sided, from stats package) or Student's t test (t.test() R function, two-sided, from stats package), depending on whether distributions were expected to be approximately normal. Comparisons of the proportion of patients with truncating alterations in PBRM1 by clinical response group were done with Fisher's exact tests when comparing CB and NCB (fisher. test() R function, two-sided, from stats package) and Fisher-Freeman-Hal ton Exact tests when comparing CB, IB, and NCB (fisher.test() R function with 2x3 contingency table, two-sided, from stats package).

Kaplan-Meier analyses were done using the R packages survival and survminer.

Significance testing for differences in progression-free survival or overall survival were calculated using the log-rank test. All comparisons were two-sided with an alpha level of 0.05. MutSig2CV was used to identify genes of interest among all those mutated in the discovery cohort. Subsequently, the Benjamini-Hochberg method for controlling false discovery rate (FDR) was applied to control for multiple hypothesis testing among the seven genes of interest with a threshold of q<0.1. All statistical analyses and figures were generated in R version 3.3.2.

Immune checkpoint inhibitors, such as nivolumab, extend the survival of a subset of patients with metastatic ccRCC (Motzer et al. (2015) N. Engl. J. Med. 373 : 1803-1813). Whether specific genomic features of ccRCC are associated with clinical benefit is unclear. In contrast to other human tumor types that respond to immunotherapy, such as non-small cell lung cancer (NSCLC), melanoma, and microsatellite-unstable colorectal

adenocarcinoma, ccRCC harbors a low burden of somatic mutations (Snyder et al. (2014) N. Engl. J. Med. 371 :2189-2199; Rizvi et al. (2015) Science 348: 124-128; Le et al. (2015) N. Engl. J. Med. 372:2509-2520; Van Allen et al. (2015) Science 350:207-211). Melanoma and NSCLC typically harbor 10 to 400 mutations per megabase (Mb) and these genetic variants can generate tumor-specific antigens (neoantigens) that stimulate a strong antitumor immune response (Motzer et al. (2015) N. Engl. J. Med. 373 : 1803-1813; Snyder et al (2014) N. Engl. J. Med. 371 :2189-2199; Rizvi et al. (2015) Science 348: 124-128; Le et al. (2015) N. Engl. J. Med. 372:2509-2520). In contrast, ccRCC harbors an average of only 1.1 mutations/Mb (Cancer Genome Atlas Research (2013) Nature 499:43-49; de Velasco et al. (2016) Cancer Immunol. Res. 4:820-822), yet it ranks highly among tumor types in terms of immune cytolytic activity (Rooney et al. (2015) Cell 160:48-615), immune infiltration score, and T cell infiltration score in the tumor microenvironment (§enbabaoglu et al. (2016) Genome Biol. 17:231).

It was hypothesized that distinct molecular mechanisms underlie the

immunologically active tumor microenvironment and responsiveness to immune checkpoint therapy in patients with ccRCC. As part of a prospective clinical trial (Choueiri et al.

(2016) Clin. Cancer Res. 22:5461-5471), pre-treatment tumors from 35 patients with metastatic ccRCC on a clinical trial of anti-programmed cell death- 1 receptor (anti- PD-1) therapy (nivolumab) were analyzed. Whole exome sequencing (WES) from paired tumor/normal tissue was performed to identify genetic correlates of clinical benefit. To validate the findings, an independent cohort of 63 patients with metastatic ccRCC treated with therapies blocking PD-1 {e.g., nivolumab) or its ligand, PD-L1 {e.g., atezolizumab), were analyzed (Figure 13 A and Table 6A).

Baseline clinical and demographic features in the discovery cohort have been previously described, and the subset of patients with complete pre-treatment molecular profiling did not differ substantially in clinical or demographic features from patients whose data did not pass technical quality control (Figures 14A-14B) or from the larger published cohort (Choueiri et al. (2016) Clin. Cancer Res. 22:5461-5471). Given previous evidence suggesting that refined clinical stratifications are necessary to assess clinical benefit from immune checkpoint blockade (Wolchok et al. (2009) Clin. Cancer Res. 15:7412-7420), a composite response endpoint incorporating RECIST (Response Evaluation Criteria In Solid Tumors) (Eisenhauer et al. (2009) Eur. J. Cancer 45:228-247), radiographic tumor shrinkage, and progression-free survival (PFS), was defined (Figure 13B and Table 6B). Clinical benefit (CB) included patients with complete response (CR) or partial response (PR) by RECIST 1.1 (i.e., tumor shrinkage >30% from baseline) (Eisenhauer et al (2009) Eur. J. Cancer 45:228-247) or stable disease (SD) if they had any objective reduction in tumor burden lasting at least 6 months. This modification to include some patients with SD is intended to differentiate those patients with naturally indolent disease (i.e., slow tumor growth not surpassing 20% of baseline tumor size) from those with tumor response to immune checkpoint inhibitors (Gofrit et al. (2015) Springer Plus 4:580). No clinical benefit (NCB) patients experienced progressive disease (PD) by RECIST 1.1 and were discontinued from immunotherapy within three months. All other patients were termed "intermediate benefit" (IB). One patient in the discovery cohort was classified as CB despite PFS < 6 months because there was continued tumor shrinkage (-67% of baseline tumor size) after an initial period of minor tumor progression, and the patient had overall survival exceeding 32 months (Figures 15A-15B). Consistent with prior observations (Motzer et al. (2015) N. Engl. J. Med. 373 : 1803-1813), the dose of nivolumab, patient gender, and baseline PD-Ll immunohistochemical staining from metastatic biopsies did not predict patient overall survival (OS) following initiation of anti-PD-1 therapy (p>0.05 for all; log-rank test) (Figure 16).

Mean exome-wide target coverage in the discovery cohort was 128-fold for tumor sequencing and 91 -fold for matched germline sequencing (Tables 6A and 6E). Overall, nonsynonymous mutation burden was moderate in the discovery cohort (median 82 per exome, range 45-157). The tumors of patients with CB and those with NCB showed similar mutation burdens and intratumoral heterogeneity (Figures 13C-13D and Table 6C). Mutations and copy number alterations affecting antigen presentation machinery and HLA class I alleles were uncommon and were present in tumors of both CB and NCB patients (Figures 17A-17B).

The analyses were next focused on the mutations most likely to be functionally important. MutSig2CV (Lawrence et al. (2013) Nature 499:214-218) was applied to identify genes recurrently mutated in the discovery cohort. Of these genes, the search was limited to highly deleterious variants, meaning known hotspot or putative truncating (frameshift insertion or deletion, nonsense mutation, or splice-site) mutations. Of the seven recurrently mutated genes (Figure 18 A) (Cancer Genome Atlas Research (2013) Nature 499:43-49), PBRM1 was the only gene in which truncating, or loss-of-function (LOF), mutations were enriched in tumors from patients in the CB vs. NCB group (9/11 vs. 3/13; Fisher's exact p=0.012, q=0.086, odds ratio for CB=12.93, 95% C.I. 1.54-190.8) (Figure 18B and Table 6D). In this cohort, all truncating PBRM1 alterations co-occurred with deletion of the non-mutated allele on chromosome 3p (Figure 18A), resulting in complete LOF of PBRMl, and most of the mutations were predicted to be clonal

(present in all tumor cells) (Table 6D). Prior large-scale sequencing studies have shown that PBRMl LOF alterations occur in up to 41% of ccRCC tumors (Varela et al. (2011) Nature 469:539-542) and are commonly clonal events present in all or nearly all tumor cells (Gerlinger et al. (2014) Nat. Genet. 46:225-233). Patients whose tumors showed biallelic PBRMl loss had significantly prolonged OS and PFS compared to patients without PBRMl LOF (log-rank p=0.0074 and p=0.029, respectively) (Figures 18C and 19), and they experienced sustained reductions in tumor burden (Figure 18D).

To evaluate the reproducibility of this finding, matched pre-treatment tumor and germline genomic data were examined from an additional 63 patients treated with anti-PD- (L)l therapy, either alone or in combination with anti-CTLA-4 therapy. Of these 63 patients, PBRMl mutation status was derived from WES in 49 patients and panel sequencing in 14 patients (Figures 20A-20B and Tables 6E-6F). Tumors from CB patients were more likely to harbor truncating alterations in PBRMl (17/27 vs. 4/19, Fisher's exact p=0.0071, odds ratio for CB=6.10, 95% C.I. 1.42-32.64) (Figures 20C-20D and Table 6G). Although copy number alterations in all samples in the validation cohort could not be assessed, it is believed that the PBRMl LOF mutations represented biallelic loss, as chromosome 3p deletions are nearly ubiquitous in ccRCC (Cancer Genome Atlas Research (2013) Nature 499:43-49). Notably, one of the four NCB patients whose tumor showed a PBRMl LOF mutation also had an alteration in B2M, which codes for a protein important in antigen presentation. This provides a potential explanation for the patient's lack of clinical benefit from immune checkpoint blockade therapy despite having a truncating PBRMl mutation.

While primary analyses excluded patients with intermediate benefit (IB) due to the unclear effect of immune checkpoint blockade therapy on patient outcomes in this group, the observed trend between PBRMl mutation status and clinical benefit persisted with the inclusion of these patients as an intermediate phenotype. In both the discovery and validation cohorts, patients in the IB group had intermediate rates of PBRMl LOF (82%, 64%, 23% for CB, IB, NCB in the discovery cohort and 63%, 41%, 21% for CB, IB, NCB in the validation cohort; Fisher-Freeman-Hal ton Exact p = 0.017 and 0.017). Additionally, while no difference in clinical benefit was observed between treatment-naive and previously-treated patients in the discovery cohort (Figures 15A-15B), the progression-free survival benefit conferred by PBRM1 LOF was more prominent in tumors from previously- treated patients compared to those from patients receiving anti-PD-1 therapy as their first cancer therapy (p=0.009) (Figure 21 and Table 6).

The PBRM1 gene codes for BAF180, a subunit of the PBAF subtype of the

SWI/S F chromatin remodeling complex. The PBAF complex suppresses the hypoxia transcriptional signature in VHL-I- ccRCC (Nargund et al. (2017) Cell Reports 18:2893- 2906; Gao et al. (2017) Proc. Natl. Acad. Sci. USA 114: 1027-1032), but its effects on tumor-immune interactions have not been thoroughly studied. To explore the potential impact of this complex on the immunophenotype of ccRCC, previously reported whole transcriptome sequencing (RNA-seq) data from A704 ccRCC cell lines with perturbations in the PBAF complex (Gao et al. (2017) Proc. Natl. Acad. Sci. USA 114: 1027-103279) were analyzed. Loss of BAF180 or the related PBAF subunit BRG1, encoded by the gene SMARCA4, prevent formation of the intact PBAF complex (Gao et al. (2017) Proc. Natl. Acad. Sci. USA 114: 1027-1032). Gene expression analyses of BAF180-null

(A704BAF 180-/-) cell lines vs. PB AF-wildtype (A704BAF 180wt) cell lines were performed and gene expression analyses of BRGl-null (A704BAF180wt, BRG1- /-) cell lines vs. PB AF-wildtype (A704BAF180wt) cell lines were also performed (Figure 22A). Differential gene expression analysis showed substantial overlaps (-50%) between the top 100 genes differentially expressed in A704BAF180-/- vs. A704BAF180wt and A704BAF180wt, BRG1-/- vs. A704BAF180wt (Table 61). This reflects the fact that BAF180 is essential to the PBAF but not the BAF complex, while BRG1 is a required subunit of both. Thus, the BAF180-null and BRGl-null cell lines have some shared characteristics but are also biologically and phenotypically distinct. Gene set enrichment analysis (GSEA) on 50 "hallmark" gene sets representing major biological processes (Subramanian et al. (2005) Proc. Natl. Acad. Sci. USA 102: 15545-15550) revealed five gene sets whose expression was significantly enriched in cell lines that were PBAF- deficient. These included genes linked to IL6/JAK-STAT3 signaling, T F-a signaling via F-κΒ, and IL2/STAT5 signaling (Figure 22A and Tables 6J-6K). As expected, the hallmark hypoxia gene set was upregulated in A704BAF180-/- vs. A704BAF180wt cell lines (family-wise error rate - FWER q=0.071) (Table 6J) (Gao et al. (2017) Proc. Natl. Acad. Sci. USA 114: 1027-1032). Across the more refined "founder" gene sets describing these five significantly enriched hallmark gene sets, the most strongly enriched gene set in PBAF-deficient cell lines was the KEGG cytokine-cytokine receptor interaction gene set (FWER q=0.0020 for A704BAF180-/- vs. A704BAF180wt and q=0.023 for A704BAF180wt, BRG1-/- vs. A704BAF180wt) (Figure 22A and Tables 6L-6U). This gene set includes both immune- stimulatory {e.g., IL12, CCL21) and immune-inhibitory {e.g., IL10) genes, but Gene Ontology term analysis showed that the genes most strongly enriched in PBAF-deficient cell lines were immune-stimulatory (Table 6V). Previously reported GSEA analysis of untreated ccRCC from The Cancer Genome Atlas (TCGA) and a murine model of PBRM1 loss also show amplified transcriptional outputs of HIF1 and STAT3, involved in hypoxia response and JAKSTAT signaling respectively, in PBRM1- mutant vs. ,P£RM7-wildtype states (Nargund et al. (2017) Cell Reports 18:2893-2906). GSEA analysis of RNA-seq from pre-treatment tumors in the discovery and validation cohorts of this study (n = 18 PBRM1 -LOF vs. n = 14 ^SRMy-intact) confirmed increased expression of the hypoxia and IL6/JAK-STAT3 gene sets in the PBRMl-LOF tumors (Figure 22B and Tables 6W-6X). Given JAK-STAT3 pathway gene involvement in the interferon gamma (IFN-g-) signaling pathway and IFN-g-dependent cancer

immunostimulation (Sharma et al. (2017) Cell 168:707-723), differential expression of these genes may impact PBRM1 -LOF patients' response to anti-PD-(L)l therapy. In addition to assessing tumor-intrinsic gene expression with GSEA, the quality of the tumor- immune microenvironment in PBRMl-LOF vs. ^SRMy-intact ccRCC was further characterized in three independent cohorts: TCGA (Cancer Genome Atlas Research (2013) Nature 499:43-49), an independent cohort of untreated ccRCC tumors (Sato) (Sato et al. (2013) Nat. Genet. 45:860-867), and patient tumors. In all three cohorts, tumors harboring LOF mutations m PBRMl showed lower expression of immune inhibitory ligands {e.g., CD276 and BTLA) (Ramsay (2013) Br. J. Haematol. 116:313-325) than those without PBRM1 mutations. This finding was unexpected as high PD-L1 staining is associated with increased responsiveness to anti-PD-1 and anti-PD-Ll agents in other cancer types

(Rosenberg et al. (2016) Lancet 387: 1909-1920; Topalian et al. (2012) N. Engl. J. Med. 366:2443-2454) and despite the fact that these differences were relatively small and in the context of differing degrees of tumor-stromal admixture (Figures 23A-23C) (§enbabaoglu et al. (2016) Genome Biol. 17:231). LOF mutations in VHL, the most commonly mutated gene in the TCGA ccRCC cohort, were also examined. VHL mutation status did not correlate with immune related gene expression (Figure 24), indicating that observed differences in immune gene expression in the context of PBRM1 LOF is believed to be specific to the PBRM1 gene. Based on the foregoing, it has been shown that patients with metastatic ccRCC harboring truncating mutations in PBRM1 experienced increased clinical benefit from immune checkpoint therapy. It is believed that this is due to distinct immune-related gene expression profiles in PBRM1 -mutant or PBAF-deficient tumor cells compared to their PBAF-intact counterparts, as shown by RNA-seq analyses described herein. In vivo studies of mice harboring tumor clones with inactivation of PBRM1 - or the related essential PBAF complex components ARID2 or BRD7 - show that cells with PBAF loss are more sensitive to T-cell-mediated cytotoxicity compared to their PBAF-intact counterparts (Pan et al. (2018) Science, in press), which helps to explain the conflicting results regarding PBRM1 mutation status as a prognostic variable in ccRCC (in the absence of immunotherapy) in prior studies (Beuselinck et al. (2015) Clin. Cancer Res. 21 : 1329-1339; Fay et al. (2016) J. Natl. Compr. Cane. Netw. 14:820-824; Hakimi et al. (2013) Clin. Cancer Res. 19:3259- 3267; Hsieh et al. (2017) Eur. Urol. 71 :405-414; Kapur et al. (2013) Lancet Oncol. 14: 159- 167; Kwiatkowski et al. (2016) Clin. Cancer Res. 22:2445-2452; Nam et al. (2015) Urol. Oncol. 33 :340.e349-316; Pawlowski et al. (2013) Int. J. Cancer 132:E11-E17; Uhlen et al. (2017) Science 357:pii eaan2507). PBRMl also previously has been linked to longer PFS with VEGF -targeted therapies (Carlo et al. (2017) Kidney Cancer 1 :49-56). Additional in vivo studies can be used to further confirm the results described herein. Given the high prevalence of PBRMl LOF in ccRCC and of SWI/SNF alterations across all cancer types (more than 20%) (Kadoch et al. (2013) Nat. Genet. 45:592-601), these results have important implications as a molecular tool for considering immunotherapy-responsiveness in ccRCC and across cancer types.

Table 6A: Whole exome sequencing metrics and inclusions/exclusions for patients in the discovery cohort

RCC 84 166.18581 88.436816 whole exome illumina coding vl 0.45 1.93 0 0

RCC 85 43.586957 168.436641 whole exome illumina coding vl 0.17 4.08 1 0

RCC 114 135.707278 77.721511 whole exome illumina coding vl 0.29 1.97 0 0

RCC 117 173.22159 68.856331 whole exome illumina coding vl 0.33 1.87 0 0

RCC 15 143.012126 79.906338 whole exome illumina coding vl 0.69 1.97 0 0

RCC 68 107.126976 88.452741 whole exome illumina coding vl 0.3 1.86 0 0

RCC 1 157.143939 89.39856 whole exome illumina coding vl 0.2 3.6 1 0

RCC 106 176.007671 81.059438 whole exome illumina coding vl 0.35 1.9 0 0

RCC 18 139.328276 75.654059 whole exome illumina coding vl 0.21 2.3 0 0

RCC 21 178.624687 105.356301 whole exome illumina coding vl 0.51 3.39 1 0

RCC 41 138.664874 93.93237 whole exome illumina coding vl 0.19 4.28 1 0

RCC 50 162.205322 85.879444 whole exome illumina coding vl 0.31 1.81 0 0

RCC 73 158.127987 100.10628 whole exome illumina coding vl 0.6 1.83 0 0

RCC 39 147.571574 114.169462 whole exome illumina coding vl 0.13 1.92 0 0

RCC 99 34.101887 117.822339 whole exome illumina coding vl 0.36 2.77 1 0

RCC 105 152.057615 91.424807 whole exome illumina coding vl 0.48 2.06 0 0

RCC 119 26.875509 90.734659 whole exome illumina coding vl 0.49 3.08 1 0

RCC 27 125.149722 97.245404 whole exome illumina coding vl 0.34 1.93 0 0

RCC 52 131.064027 90.415506 whole exome illumina coding vl 0.54 1.88 0 0

RCC 97 210.012354 98.486524 whole exome illumina coding vl 0.38 2.2 0 0

RCC 2 159.912441 69.844188 whole exome illumina coding v 1 0.52 1.68 0 DeathUnrelate dCancer

RCC 72 9.627872 94.01896 whole exome illumina coding vl NA NA NA LowCoverage

RCC 5 8.689284 89.713424 whole exome illumina coding vl 0.36 1.98 0 LowCoverage

RCC 54 0.007939 84.883698 whole exome illumina coding vl NA NA NA LowCoverage

RCC 100 7.711684 105.962605 whole exome illumina coding vl 0.34 2.01 0 LowCoverage

RCC 47 0.298156 95.4427 whole exome illumina coding vl NA NA NA LowCoverage

RCC 66 8.71954 98.033649 whole exome illumina coding vl 0.46 2.16 0 LowCoverage

RCC 43 105.603458 72.354112 whole exome illumina coding vl 0.06 2.43 0 LowPurit

RCC 12 162.560923 104.266666 whole exome illumina coding vl 0.05 2.74 0 LowPurity

RCC 24 166.047506 75.247762 whole exome illumina coding vl 0.1 2.46 0 LowPurity

RCC 40 154.736269 87.045058 whole exome illumina coding vl 0.1 2.44 0 LowPurity

RCC 8 154.801856 83.048353 whole exome illumina coding vl NA NA NA LowPurity

RCC 103 138.626523 96.365324 whole exome illumina coding vl LowPurity

RCC 26 159.566974 100.887491 whole exome illumina coding vl 0.07 2.96 0 LowPurity

RCC 95 143.956046 90.060356 whole exome illumina coding vl 0.09 2.57 0 LowPurity

RCC 17 129.343681 81.980679 whole exome illumina coding vl 0.04 3.61 1 LowPurity

RCC 22 144.076612 97.672268 whole exome illumina coding vl 0.06 2.91 0 LowPurity

RCC 28 162.443009 89.968028 whole exome illumina coding vl 0.08 2.45 0 LowPurity

RCC 29 150.205436 89.123637 whole exome illumina coding vl NA NA NA LowPurity

RCC 6 145.806274 83.646769 whole exome illumina coding vl 0.07 2.69 0 LowPurity

RCC 45 132.158193 79.179771 whole exome illumina coding vl 0.06 2.58 0 LowPurity

Table 6B: Clinical characteristics of patients receiving anti-PDl therapy (nivolumab) in discovery cohort (N=35) (All patients at wes of 1, nivolumab as drug)

RCC MALE 64 10 0 PD 17 11 0 45 0 100 0 0 0 0 0 no 105 mg/kg 8 clinical benefit

RCC MALE 72 0.3 0 PD 13 30 0 37 0 100 0 0 0 0 0 no

39 mg kg 6 clinical benefit

RCC MALE 77 2 0 PD 11 17 0 58 0 100 0 0 0 0 0 no

73 mg kg 8 clinical benefit

RCC FEMA 63 10 1 SD -67 98 1 86 0 91 5 3 1 1 1 clinical 50 LE mg/kg 2 benefit

-N

RCC MALE 66 10 0 PD 10 19 1 42 0 100 0 0 0 0 0 no 41 mg/kg 5 clinical benefit

RCC MALE 68 0.3 0 PD 65 11 1 43 0 NA NA NA NA NA NA no 18 mg/kg 1 clinical benefit

RCC MALE 49 10 0 PD 24 43 1 38 0 100 0 0 0 0 0 no 1 mg/kg clinical benefit

RCC MALE 69 0.3 0 PD -4 74 0 39 0 100 0 0 0 0 0 no 68 mg/kg 0 clinical benefit

RCC FEMA 73 10 1 SD -10 10 0 66 0 99 1 0 0 0 0 clinical 15 LE mg/kg 13 3 benefit

-N

RCC FEMA 57 0.3 0 PR -51 34 1 20 0 100 0 0 0 0 0 clinical 114 LE mg/kg 0 8 benefit

RCC FEMA 55 0.3 0 SD 8 68 0 88 0 95 5 0 0 0 0 interme 84 LE mg/kg 0 diate benefit

RCC MALE 64 0.3 0 SD 3 16 1 10 0 100 0 0 0 0 0 interme 102 mg/kg 5 8 diate benefit

RCC FEMA 75 2 0 PD 12 58 1 37 0 100 0 0 0 0 0 no

96 LE mg/kg 2 clinical benefit

RCC MALE 54 0.3 0 PD 57 49 0 36 0 25 35 20 20 1 1 no 90 mg/kg 9 clinical benefit

RCC MALE 60 2 0 SD 15 36 1 87 0 100 0 0 0 0 0 interme 115 mg/kg 6 diate benefit

RCC MALE 64 10 0 PR -43 68 0 50 0 100 0 0 0 0 0 clinical

93 mg/kg 4 0 benefit

RCC FEMA 61 2 0 SD 9 87 1 13 0 100 0 0 0 0 0 interme

79 LE mg/kg 3 0 diate benefit

RCC FEMA 62 2 0 SD 17 99 1 81 0 100 0 0 0 0 0 interme 56 LE mg/kg 2 diate benefit

RCC FEMA 72 10 1 PD 16 16 0 39 0 100 0 0 0 0 0 no 25 LE mg/kg 7 clinical

-N benefit

RCC MALE 59 10 1 PR -86 10 1 54 0 100 0 0 0 0 0 clinical 14 mg/kg 25 1 benefit

-N

RCC MALE 50 10 1 SD 9 10 1 12 0 100 0 0 0 0 0 interme 11 mg/kg 24 2 diate

-N benefit

RCC FEMA 48 10 0 SD 7 10 1 87 0 NA NA NA NA NA NA interme

62 LE mg/kg 6 diate benefit

RCC MALE 64 10 0 SD -13 34 1 43 1 96 4 0 0 1 0 interme 21 mg/kg 9 diate benefit

RCC FEMA 61 2 0 PR -61 87 1 82 1 100 0 0 0 0 0 clinical 106 LE mg/kg 0 1 benefit

RCC MALE 58 10 1 SD -10 64 1 38 1 98 2 0 0 0 0 interme 117 mg/kg diate

-N benefit RCC MALE 36 2 0 SD -12 71 0 42 1 95 2 2 1 1 0 interme 85 mg kg 2 diate benefit

RCC FEMA 55 2 0 PR -50 98 1 68 1 20 25 25 30 1 1 clinical 58 LE mg/kg 8 7 benefit

RCC FEMA 64 10 1 SD 0 10 1 17 1 97 3 0 0 1 0 interme 10 LE mg/kg 58 3 diate

-N benefit

RCC MALE 65 10 1 SD -13 24 0 21 1 100 0 0 0 0 0 clinical

32 mg/kg 0 3 benefit

-N

RCC MALE 76 10 1 CR 10 1 10 1 100 0 0 0 0 0 clinical 20 mg/kg 10 65 22 benefit

-N 0

RCC MALE 50 2 0 PD NA 41 0 41 0 97 3 0 0 1 0 no

99 mg/kg clinical benefit

Table 6C. Summary of mutational burden in the discovery cohort (N=35)

RCC 102 73 29 5 87 5 10 65 4 4 41

RCC 171 126 45 10 122 31 18 92 20 14 50

RCC 70 45 25 0 36 29 5 22 23 0

73

RCC 126 86 40 0 117 2 7 81 2 3

39

RCC 155 102 53 6 64 57 34 49 48 5

99

RCC 152 105 47 5 118 18 16 87 15 3 105

RCC 97 62 35 4 81 2 14 60 2 0 119

RCC 128 98 30 8 78 41 9 63 32 3

27

RCC 125 94 31 8 94 15 16 77 13 4 52

RCC 116 82 34 5 77 26 13 62 20 0

97

Table 6D: Truncating PBRMl alterations in patients discovery cohort passing whole exome quality control (N=35)

R NA 221.56 NA NA NA NA NA NA N N N N N

C A A A A A

C

5

6

R PB 67.19 3 526 52 Fram T T - p.Nl DE 0.4 13 15 1 str

C RM 214 62 e Shi 017fs L 64 elk

C 1 87 14 ft De 28 a,

7 87 1 57 ind

9 14 elo cat or

R PB 173.78 3 526 52 Splic C c T SN 0.1 6 41 1 N

C RM 512 65 e_Sit P 27 A

C 1 77 12 e 66

9 77

3

R NA 115.51 NA NA NA NA NA NA N N N N N

C A A A A A

C

1

15

R NA 124.87 NA NA NA NA NA NA N N N N N

C A A A A A

C

9

0

R PB 164.39 3 526 52 Fram A A - p.S81 DE 0.4 91 13 1 str

C RM 434 64 e Shi 8fs L 02 5 elk

C 1 89 34 ft De 65 a,

9 89 1 48 ind

6 67 elo cat or

R PB 266.4 3 526 52 Splic C C T SN 0.2 25 81 1 N

C RM 630 66 e_Sit P 35 A

C 1 52 30 e 84

1 52 9

02

R NA 95.31 NA NA NA NA NA NA N N N N N

C A A A A A

C

5

8

R PB 130.86 3 526 52 Fram T T - p.Kl DE 0.1 12 58 1 str

C RM 962 69 e Shi 35fs L 71 elk

C 1 72 62 ft De 42 a,

8 72 1 85 ind

4 71 elo cat or

R NA 47.52 NA NA NA NA NA NA NA NA NA N N N N N

C A A A A A

C

Q o

5

R PB 111.22 3 526 52 Fram A A - p.N4 DE 0.1 10 82 1 str

C RM 629 66 e Shi 63fs L 08 elk

C 1 64 29 ft De 69 a,

1 64 1 56 ind

14 52 elo cat or

R PB 146.69 3 526 52 Nons G G A p.Q8 SN 0.2 36 89 1 N

C RM 433 64 ense_ 09* P 88 A

C 1 75 33 Mutat

1 75 ion

17

R PB 94.84 3 526 52 Nons C C A p.Ell SN 0.5 53 47 1 N

C RM 131 61 ense_ 05* P 3 A

C 1 94 31 Mutat

1 94 ion

5 R NA 100.73 NA NA NA NA NA NA N N N N N

C A A A A A

C

6

8

R NA 138.81 NA NA NA NA NA NA N N N N N

C A A A A A

C

1

R PB 155.18 3 526 52 Fram ATTTT ATTTT - p.KIl DE 0.0 19 26 0 str

C RM 206 62 e Shi 087fs L 67 4 elk

C 1 10 06 ft De 13 a,

1 14 1 78 ind

06 09 elo cat or

R PB 126.07 3 526 52 Nons C c A p.E29 SN 0.1 8 45 1 N

C RM 787 67 ense_ 1* P 50 A

C 1 48 87 Mutat 94

1 48 ion 3

8

R PB 125.64 3 526 52 Fram T T - p.Kl DE 0.4 53 67 1 str

C RM 132 61 e Shi 146fs L 41 elk

C 1 10 32 ft De 66 a,

2 10 1 66 ind

1 67 elo cat or

R NA 123.22 NA NA NA NA NA NA N N N N N

C A A A A A

C

4

1

R PB 135.79 3 527 52 Splic c c T SN 0.2 22 81 1 N

C RM 125 71 e_Sit P 13 A

C 1 15 25 e 59

5 15 2

0

R NA 181.65 NA NA NA NA NA NA N N N N N

C A A A A A

C

7

3

R NA 130.38 NA NA NA NA NA NA N N N N N

C A A A A A

C

3

9

R NA 28.13 NA NA NA NA NA NA N N N N N

C A A A A A

C

9

9

R NA 146.47 NA NA NA NA NA NA N N N N N

C A A A A A

C

1

05

R PB 28.57 3 526 52 Splic c c G SN 0.6 10 5 1 N

C RM 824 68 e_Sit P 66 A

C 1 59 24 e 66

1 59 7

19

R PB 248.99 3 525 52 In Fr TCATC TCATC - p.AK DE 0.1 14 77 1 ind

C RM 980 59 ame ATCTA ATCTA WD L 5 elo

C 1 81 81 Del CCACT CCACT DE12 cat

2 01 TTAGC TTAGC 49del or

7 A A

R PB 97.9 3 526 52 Fram T T - p.Dl DE 0.2 21 68 1 str

C RM 132 61 e Shi 148fs L 35 elk

C 1 05 32 ft De 95 a,

5 05 1 50 ind

2 56 elo cat or

R PB 119.39 3 526 52 Nons C C A p.E41 SN 0.2 22 57 1 N C RM 630 66 ense_ 7* P 78 A C 1 08 30 Mutat 48

9 08 ion 1

7

Table 6E: Sequencing Metrics and Inclusion/Exclusion Criteria for Whole Exome

Sequencing in Validation Cohort (N=67)

RCC- WES 70.42 82.15 NA whole exome illumina 0.45 1.65 0 0 IM 002 7352 0509 coding vl

RCC- WES 160.0 172.7 NA whole exome illumina NA NA NA NonCle

ΓΜ 003 6141 0407 coding vl arCellR

2 4 CC

RCC 2810 WES 80.37 89.49 NA whole exome illumina 0.44 1.89 0 0 66 1008 6933 coding vl

RCC 3719 WES 93.84 91.06 NA whole exome illumina 0.48 2.09 0 0 82 6727 197 coding vl

RCC 3903 WES 78.22 116.3 NA whole exome illumina 0.46 1.92 0 0 92 7866 0660 coding vl

1

RCC 4708 WES 179.5 92.16 NA whole exome illumina 0.45 1.77 0 0 74 4567 1197 coding vl

1

RCC 4727 WES 97.99 81.59 NA whole exome illumina 0.39 1.86 0 0 70 001 2339 coding vl

RCC 5046 WES 182.3 116.9 NA whole exome illumina 0.47 2.05 0 0 42 1248 8203 coding vl

4

RCC 5092 WES 129.1 114.1 NA whole exome illumina 0.57 1.89 0 0 14 4603 8974 coding vl

9 6

RCC 5197 WES 159.5 69.79 NA whole exome illumina 0.78 1.91 0 0 4086 1169 7316 coding vl

2

RCC 5546 WES 223.8 93.01 NA whole exome illumina 0.39 2.92 1 0 52 5971 6058 coding vl

7

MCA1 targete NA NA NA Caris Molecular NA NA NA 0 d_pane Intelligence+600 gene

1 NGS-2015

MCA2 targete NA NA NA FoundationOne- (315 NA NA NA 0 d_pane genes, 28 introns) 2014

1

MCA3 targete NA NA NA Caris Molecular NA NA NA 0 d_pane Intelligence+600 gene

1 NGS-2015

MCA4 targete NA NA NA Caris Molecular NA NA NA 0 d_pane Intelligence+600 gene

1 NGS-2016

MCA5 targete NA NA NA FoundationOne- (236 NA NA NA 0 d_pane genes, 47 introns) 2014

1

MCA6 targete NA NA NA Caris Molecular NA NA NA 0 d_pane Intelligence+600 gene

1 NGS-2016

MCA7 targete NA NA NA FoundationOne- (236 NA NA NA 0 d_pane genes, 47 introns) 2014

1

MCA8 targete NA NA NA Caris Molecular NA NA NA 0 d_pane Intelligence+600 gene

1 NGS-2016

MCA9 targete NA NA NA Caris Molecular NA NA NA 0 d_pane Intelligence+600 gene

1 NGS-2017

MCA10 targete NA NA NA Caris Molecular NA NA NA 0 d_pane Intelligence+600 gene

1 NGS-2016

MCA11 targete NA NA NA FoundationOne- (315 NA NA NA 0 d_pane genes, 28 introns) 2016

1

MCA12 targete NA NA NA Caris Molecular NA NA NA 0 d_pane Intelligence+600 gene

1 NGS-2015

MCA13 targete NA NA NA FoundationOne- (236 NA NA NA 0 d_pane genes, 47 introns) 2014

1

MCA14 targete NA NA NA Caris Molecular NA NA NA 0 d_pane Intelligence+600 gene

1 NGS-2016 Table 6F: Clinical information for immune checkpoint-treated patients in validation cohort (N=63)

PGDX2818T nivolu CR MAL 5 0 -100 clear 253 1 2533 1 clinical

Ex- mab ¹ E 1 -cell 3 benefit RCC032PT1

PGDX2817T nivolu PD MAL 4 0 42 clear 892 0 112 0 intermedi

Ex- mab ¹ E 3 -cell ate benefit RCC031PT1

PGDX2816T nivolu PR MAL 5 0 -71 clear 175 1 1124 0

¹ clinical

Ex- mab E 8 -cell 5 benefit RCC030PT1

PGDX2815T nivolu PD FEM 3 0 52 clear 501 0 59 0 no clinical

Ex- mab ¹ ALE 5 -cell benefit RCC029PT1

PGDX2814T nivolu PD FEM 6 0 33 clear 148 0 51 0 no clinical

Ex- mab ¹ ALE 7 -cell benefit RCC028PT1

PGDX2813T nivolu CR MAL 6 0 -100 clear 220 0 2012 0 clinical

Ex- mab ¹ E 8 -cell 8 benefit RCC027PT1

PGDX2811T nivolu PR MAL 7 0 -90 clear 281 1 2810 1 clinical

Ex- mab ¹ E 3 -cell 0 benefit RCC025PT1

BL5166 T1 nivolu SD MAL 6 0 -11 clear 622 0 156 0 intermedi mab ¹ E 3 -cell ate benefit

RENAL- nivolu PR MAL 6 0 -37 clear 499 1 499 1 clinical 15349 CCP mab ¹ E 7 -cell benefit M 0600855

RENAL- nivolu PD MAL 7 0 NA clear 293 1 66 0 no clinical 15349 CCP mab ¹ E 3 -cell benefit M 0600862

RCC- nivolu PR NA 4 0 -39 clear 379 0 280 0 clinical Π 001 mab ¹ 0 -cell benefit

RCC- nivolu SD NA 6 0 -16 clear NA NA 735 0 clinical

Π 002 mab ¹ 1 -cell benefit

RCC 281066 nivolu SD M 6 0 NA clear 460 1 460 1 intermedi mab ¹ 0 -cell ate benefit

RCC 371982 nivolu PD M 7 0 NA clear 448 0 71 0 no clinical mab ¹ 0 -cell benefit

RCC 390392 nivolu SD M 7 0 2 clear 174 1 55 1 intermedi mab ¹ 7 -cell ate benefit

RCC 470874 nivolu PD M 5 0 NA clear 247 0 42 0 no clinical mab ¹ 9 -cell benefit

RCC 472770 nivolu PD M 5 0 NA clear 558 0 84 0 no clinical mab ¹ 2 -cell benefit

RCC 504642 nivolu PD F 5 0 NA clear 102 0 41 0 no clinical mab ¹ 5 -cell benefit

RCC 509214 nivolu PR M 4 0 NA clear 370 0 204 0 al mab ¹ clinic

4 -cell benefit

RCC 519740 nivolu SD F 8 0 3.5 clear 456 0 220 0 intermedi 86 mab ¹ 1 -cell ate benefit

RCC 554652 nivolu PR F 7 0 NA clear 484 1 336 1

¹ clinical mab 6 -cell benefit

MCA1 atezoli 0 PD NA N 0 NA clear NA NA 85 0 no clinical zumab A -cell benefit

MCA2 atezoli 0 PD NA N 0 NA clear NA NA 83 0 no clinical zumab A -cell benefit

MCA3 atezoli 0 SD NA N 0 -15 clear NA NA 337 0 clinical zumab A -cell benefit

MCA4 nivolu 0 PD NA N 0 NA clear NA NA 145 0 intermedi mab A -cell ate benefit

MCA5 nivolu 0 PD NA N 0 NA clear NA NA 203 0 intermedi mab A -cell ate benefit

MCA6 nivolu 0 CR NA N 0 NA clear NA NA 196 0 clinical mab A -cell benefit

MCA7 nivolu 0 PR NA N 0 -50 clear NA NA 601 0 clinical mab A -cell benefit

MCA8 nivolu 0 PD NA N 0 NA clear NA NA 107 0 no clinical mab A -cell benefit

MCA9 nivolu 0 PD NA N 0 NA clear NA NA 31 0 no clinical mab A -cell benefit

MCA10 nivolu 0 SD NA N 0 NA clear NA NA 312 0 intermedi mab A -cell ate benefit

MCA11 nivolu 0 SD NA N 0 NA clear NA NA NA 0 intermedi mab A -cell ate benefit MCA12 nivolu 0 PR NA N 0 -53 clear NA NA 127 0 clinical mab A -cell benefit

MCA13 nivolu 0 PR NA N 0 -43 clear NA NA NA 0 clinical mab A -cell benefit

MCA14 nivolu 0 PD NA N 0 NA clear NA NA 65 0 no clinical mab A -cell benefit

Table 6G: Truncating PBRMl alterations in validation cohort (N=63)

PD 023 PB 3 526 526 Splice T T A NA SN 0.2 9 33 1 NA NA

RM 630 630 Site P 142

1 53 53 86

PD 024 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

PD 025 PB 3 525 525 Frame C C p.Gl DE 0.1 21 115 1 strelk NA

RM 958 958 Shift D 429f L 544 a,

1 29 29 el s 117 indel

65 ocato r

PD 026 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

RCC.PD1 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA .DNA.10

26

RCC.PD1 PB 3 525 525 Frame C C p.Al DE 0.1 43 279 1 strelk NA

.DNA.11 RM 958 958 Shift D 438f L 335 a,

01 1 04 04 el s 403 indel

73 ocato r

RCC.PD1 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA .DNA.11

37

RCC.PD1 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA .DNA.94

4

RCC.PD1 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA .DNA.94

9

VA1008 PB 3 526 526 Frame T T p.K6 DE 0.0 15 253 1 indel NA

RM 439 439 Shift D 19fs L 6 ocato

1 43 43 el r

PGDX28 PB 3 526 526 Nonsen G G T p.33 SN 0.4 30 32 NA NA NA 18T Ex- RM 523 523 se Mut 1* P 838

RCC032P 1 06 06 ation 71

Tl

PGDX28 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 17T Ex- RCC031P

Tl

PGDX28 PB 3 525 525 Frame T T NA DE 0.3 18 39 NA NA NA 16T Ex- RM 982 982 Shift D L 157

RCC030P 1 4 4 el 89

Tl

PGDX28 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 15T Ex- RCC029P

Tl

PGDX28 PB 3 526 526 Nonsen T T A p.K6 SN 0.3 19 31 NA NA NA 14T Ex- RM 189 189 se Mut 21* P 8

RCC028P 1 79 79 ation

Tl

PGDX28 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 13T Ex- RCC027P

Tl

PGDX28 PB 3 526 526 Frame T T NA DE 0.3 35 72 NA NA NA 11T Ex- RM 574 574 Shift D L 271

RCC025P 1 32 32 el 03

Tl

BL5166 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA Tl

RENAL- PB 3 526 526 Nonsen G G A p.R8 SN 0.0 7 260 0 NA NA 15349 C RM 375 375 se Mut 89* P 262

CPM 060 1 55 55 ation 17

0855

RENAL- NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 15349 C

CPM 060

0862

RCC- NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA IM 001

RCC- NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

IM 002 RCC 281 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 066

RCC 371 PB 3 526 526 Frame C C p.A2 DE 0.1 12 52 1 strelk NA 982 RM 824 824 Shift D 49fs L 875 a,

1 28 28 el indel ocato r

RCC 390 PB 3 527 527 Splice C C NA DE 0.3 21 45 1 strelk NA 392 RM 125 125 Site L 181 a,

1 15 15 818 indel

18 ocato r

RCC 470 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 874

RCC 472 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 770

RCC 504 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 642

RCC 509 PB 3 526 526 Splice C C A NA SN 0.3 60 110 1 NA NA 214 RM 629 629 Site P 529

1 09 09 41

RCC 519 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 74086

RCC 554 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA 652

MCA1 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA pos itiv e

MCA2 PB NA NA NA Splice NA NA NA NA SN NA NA NA NA NA NA

RM Site P

1

MCA3 PB NA NA NA Frame NA NA NA NA FS 0.3 NA NA NA NA neg

RM Shift 3 ativ

1 e

MCA4 PB NA NA NA Frame NA NA NA p.N6 FS 0.0 NA NA NA NA neg

RM Shift 09fs 9 ativ

1 e

MCA5 PB NA NA NA Frame NA NA NA p.Y6 FS NA NA NA NA NA neg

RM Shift 08fs ativ

1 *34 e

MCA6 PB NA NA NA Missen NA NA NA N25 SN 0.3 NA NA NA NA neg

RM se Mut 8S P 2 ativ

1 ation L e

OF

MCA7 PB NA NA NA Nonsen NA NA NA p.El SN NA NA NA NA NA NA

RM se Mut 124* P

1 ation

MCA8 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA pos itiv e

MCA9 PB NA NA NA In Fra NA NA NA p.W DE 0.2 NA NA NA NA NA

RM me Del 141 LI 1

1 Ins L14 NS

5Del

InsC

MCA10 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA pos itiv e

MCA11 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

MCA12 PB NA NA NA Frame NA NA NA p.Kl FS 0.2 NA NA NA NA neg

RM Shift 54fs 5 ativ

1 e

MCA13 PB NA NA NA Missen NA NA NA p.S6 SN NA NA NA NA NA NA

RM se Mut 81R P

1 ation

MCA14 NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA pos itiv e

Table 6H: SWI/SNF genes Hugo Symbol Other Names BAF PBAF

ACTL6A BAF53A ^{1 1}

ACTL6B BAF53B ^{1 1}

ARID2 BAF200 ¹

BCL7A ^{1 1}

BCL7B ^{1 1}

BCL7C ^{1 1}

BCL11A ^{1 1}

BCL11B ^{1 1}

BRD7 ¹

BRD9 ¹ 0

DPF1 BAF45B ¹ 0

DPF2 BAF45D ¹ 0

DPF3 BAF45C ¹ 0

PBRM1 BAF180 0 1

PHF10 BAF45A 0 1

SMARCA2 BRM ¹ 0

SMARCA4 BRG ¹ 1

SMARCB1 BAF47, SNF1, INI1 ¹ 1

SMARCC1 BAF155 ¹ 1

SMARCC2 BAF170 ¹ 1

SMARCE1 BAF57 ¹ 1

SS18 ¹ 0

SS18L1 CREST ¹ 0

SMARCD1 BAF60A ¹ 1

SMARCD2 BAF60B ¹ 1

SMARCD3 BAF60C ¹ 1

ARID 1 A BAF250A ¹ 0

ARID IB BAF250B ¹ 0

Table 61: Intersection of top 100 positively differentially expressed genes in PBRMlnull and BRGlnull, and top 100 negatively differentially expressed genes in PBRMl null and BRGl null, both with respect to wild type using EdgeR

JAGl ERAP2

NTM ACE2

SFRP4 PADI1

SDC1 SERPINE1

TFPI2 KIAA1486

NMB B3GNT3

SLC17A3 F2R

CXCL1 PKP3

RASSF2 CHSY3

HMGCS1 ACSL5

SC4MOL DOCK2

ANGPTL4 CD74

UPB1 TAGLN

PTPRD FGF5

MACROD2 ADD2

PEG10 TUBA4A

SULF2 HKDC1

KMO RP11-428C6.1

C1QL4 SPNS2

P2RY6 UNC13D

NPR3 CAPG

SCD KRTCAP3

TTYH3 SH3KBP1

MAPK12 CLTB

MAPK11 MARCH4

CD70 ABCA13

PDZD2 KRT8

RDH10 WWC1

ITM2B MT2A

OLR1 MYEOV

NPTXR ANKRD1

FAM84B QSOX1

RASSF6 SLC1A1

LGI4 CGN

TNFSF10 VCAN

FGF9 SEMA6A

NXN

CRYAB

ADAMTS7

PKDCC

MYO10 Table 6 J: GSEA for BAF180-null vs. BAF180-wildtype A704 cell lines on Hallmark gene sets

Table 6K: GSEA for BAF180-wildtype, BRGl-null vs. BAF180-wildtype, BRGl-wildtype

A704 cell lines on Hallmark gene sets

HALLMARK INTERFERON 94 0.30972 1.09 0.310385 0.565996 1566 tags=14%, list=7%, ALPHA RESPONSE 162 3943 08 5 ¹ signal=15%

6

HALLMARK BILE ACID M 105 0.30161 1.07 0.334821 0.574843 2284 tags=15%, list=10%, ETABOLISM 64 6089 43 17 ¹ signal=17%

1

HALLMARK G2M CHECKP 198 0.26063 0.99 0.490585 0.734996 7925 tags=36%, list=36%, OINT 028 9550 77 8 ¹ signal=55%

6

HALLMARK APOPTOSIS 156 0.25963 0.97 0.533475 0.758210 6508 tags=35%, list=29%,

703 3731 04 1 ¹ signal=49%

9

HALLMARK MTORC1 SIGN 197 0.26446 1.00 0.493670 0.758576 5007 tags=22%, list=23%, ALING 19 1003 88 6 ¹ signal=28%

HALLMARK UV RESPONSE 152 0.26901 1.00 0.467320 0.772115 5325 tags=29%, list=24%, UP 275 5965 26 4 ¹ signal=38%

7

HALLMARK PEROXISOME 99 0.27292 0.97 0.525539 0.773513 2300 tags=15%, list=10%,

284 7328 16 5 ¹ signal=17%

24

HALLMARK FATTY ACID 151 0.25714 0.95 0.574034 0.786094 5981 tags=28%, list=27%, METABOLISM 728 5841 33 1 ¹ signal=38%

06

HALLMARK ANDROGEN R 95 0.26842 0.94 0.571268 0.788130 5621 tags=31%, list=25%, ESPONSE 615 5029 26 64 ¹ signal=41%

2

HALLMARK HEME METAB 187 0.24242 0.91 0.663512 0.830086 5715 tags=26%, list=26%, OLISM 312 8738 1 7 ¹ signal=34%

07

HALLMARK XENOBIOTIC 183 0.23199 0.87 0.737394 0.896378 5202 tags=24%, list=23%, METABOLISM 143 9329 9 34 ¹ signal=31%

44

HALLMARK P53 PATHWAY 193 0.22611 0.85 0.790356 0.915972 4727 tags=23%, list=21%,

341 7357 4 4 ¹ signal=29%

7

HALLMARK UNFOLDED P 112 0.14780 0.53 1 0.997495 4453 tags=14%, list=20%, ROTEIN RESPONSE 428 4943 2 ¹ signal=18%

64

HALLMARK PI3K AKT MT 98 0.22171 0.78 0.859410 1 7204 tags=30%, list=32%, OR SIGNALING 06 2928 4 ¹ signal=44%

1

HALLMARK ADIPOGENESI 193 0.20610 0.77 0.910432 1 5963 tags=23%, list=27%, S 817 3433 04 ¹ signal=31%

6

HALLMARK DNA REPAIR 147 0.19458 0.71 0.947882 1 9455 tags=36%, list=43%,

589 4612 7 ¹ signal=62%

8

HALLMARK PANCREAS BE 28 0.24026 0.69 0.886092 1 3531 tags=21%, list=16%, TA CELLS 519 2461 7 ¹ signal=25%

5

HALLMARK MITOTIC SPIN 197 0.17785 0.67 0.979936 1 6069 tags=27%, list=27%, DLE 007 7390 66 ¹ signal=37%

34

HALLMARK PROTEIN SEC 95 0.16087 0.56 0.996625 1 7983 tags=32%, list=36%, RETION 638 3520 4 ¹ signal=49%

2

Table 6L: GSEA for BAF180-null vs. BAF180-wildtype A704 cell lines on Cholesterol

Homeostasis Founder gene sets

REACTOME CHOLESTEROL BIOSYNT 21 0.81049 2.36143 0 0 0 2246 tags=67 HESIS 263 85

list=10% signal=7 4%

HORTON SREBF TARGETS 25 0.76518 2.30762 0 0 0 1908 tags=52

434 8 %, list=9%, signal=5

7%

KEGG STEROID BIOSYNTHESIS 16 0.80013 2.12638 0 0 0 2462 tags=69

51 28 %, list=ll% signal=7

7%

PODAR RESPONSE TO ADAPHOSTIN 17 0.76344 2.04900 0 2.47E-04 0.001 1302 tags=53 DN 514 96 %, list=6%, signal=5 6%

WENG POR TARGETS GLOBAL UP 18 0.68691 1.92081 0.004329 0.001405 0.004 1763 tags=39

03 44 004 882 %, list=8%, signal=4 2%

WENG POR TARGETS LIVER UP 37 0.53671 1.75533 0.002164 0.010265 0.04 1763 tags=30

414 03 502 792 %, list=8%, signal=3 2%

LE EGR2 TARGETS DN 101 0.40535 1.62636 0.004694 0.025758 0.109 1862 tags=18

05 77 836 89 %, list=8%, signal=l

9%

JI RESPONSE TO FSH UP 70 0.43719 1.62489 0.004385 0.023568 0.111 2601 tags=33

202 09 965 87 %, list=12% signal=3

7%

HOXA9 DN.V1 DN 184 0.37380 1.62192 0 0.022287 0.116 2709 tags=23

037 71 892 %, list=12% signal=2 6%

BURTON ADIPOGENESIS 10 28 0.51296 1.57341 0.027777 0.027410 0.158 2601 tags=36

09 06 778 874 %, list=12% signal=4 0%

CSR LATE UP.V1 DN 156 0.34173 1.44870 0.007211 0.062628 0.332 3441 tags=29

56 41 539 604 %, list=15% signal=3 5%

GERY CEBP TARGETS 113 0.35047 1.40656 0.027210 0.075823 0.418 943 tags=12

704 94 884 3 %, list=4%, signal=l

3%

COULOUARN TEMPORAL TGFBl SIG 127 0.31372 1.28478 0.051764 0.146267 0.679 2090 tags=16 NATURE DN 902 94 704 34 %, list=9%, signal=l

7%

MTOR UP.V1 UP 152 0.29891 1.27965 0.048223 0.141891 0.69 3119 tags=24

714 76 35 39 %, list=14% signal=2

7%

ZHANG_GATA6_TARGETS_DN 62 0.32323 1.16922 0.214622 0.254302 0.893 2796 tags=24

2 38 65 23 %, list=13% signal=2 8%

UEDA PERIFERAL CLOCK 164 0.26870 1.14285 0.187179 0.276746 0.924 2961 tags=17

546 49 78 %, list=13% signal=2 0%

CHANG CORE SERUM RESPONSE DN 198 0.25709 1.12844 0.150121 0.281890 0.941 2863 tags=18

897 94 06 48 %, list=13% signal=2 0%

GUO TARGETS OF IRS 1 AND IRS2 91 0.28360 1.09607 0.275395 0.313350 0.964 1862 tags=18

868 77 04 95 %, list=8%, signal=l

9%

AKT UP.V1 UP 155 0.24117 1.02956 0.372151 0.415859 0.989 3180 tags=22

097 77 9 9 %, list=14% signal=2 5%

WENG POR DOSAGE 19 0.31340 0.88141 0.620689 0.704599 0.999 537 tags=l l

367 83 63 86 %, list=2%, signal=l 1%

Table 6M: GSEA for BAF180-wildtype, BRGl-null vs. BAF180-wildtype, BRGl-wildtype

A704 cell lines on Cholesterol homeostasis founder gene sets

signal=2

4%

UEDA PERIFERAL CLOCK 163 0.2869 1.0742 0.32627 0.48380 1 705 tags=33

2022 894 118 232 5 %, list=32

%, signal=4 8%

LE EGR2 TARGETS DN 100 0.2848 1.0112 0.48049 0.54905 1 610 tags=37

6618 084 054 51 4 %, list=27 %, signal=5 1%

AKT UP MTOR DN.VI UP 165 0.2669 0.9899 0.51160 0.56489 1 291 tags=16

0233 1686 336 664 0 %, list=13

%, signal=l

9%

AKT UP.V1 UP 156 0.2750 1.0140 0.44930 0.56910 1 428 tags=25

6077 308 63 42 8 %, list=19

%, signal=3 1%

GUO TARGETS OF IRS 1 AND IRS2 89 0.2898 1.0152 0.44724 0.59534 1 519 tags=30

395 373 77 82 2 %, list=23

%, signal=3 9%

GOTZMANN EPITHELIAL TO MESENCHY 190 0.2474 0.9271 0.63693 0.66669 1 358 tags=16 MAL TRANSITION DN 9233 7135 6 047 8 %, list=16

%, signal=l

9%

UEDA CENTRAL CLOCK 81 0.2296 0.8009 0.80622 0.85129 1 574 tags=26

8177 6096 84 535 2 %, list=26

%, signal=3 5%

WENG POR DOSAGE 19 0.2510 0.6606 0.90150 0.95465 1 270 tags=16

292 327 48 78 5 %, list=12

%, signal=l 8%

Table 6N: GSEA for BAF180-null vs. BAF180-wildtype A704 cell lines on

IL6 JAK STAT Founder gene sets

TENEDINI MEGAKARYO TENEDINI MEGAKARYO Deta 61 0.39 1.42 0.03 0.075 0.2 35 tags CYTE MARKERS CYTE MARKERS ils ... 5845 933 7946 2779 63 03 =26

68 43 9 %, list=

16%, signa

1=31

%

BIOCARTA IL10 PATHW BIOCARTA IL10 PATHW Deta 16 0.53 1.43 0.08 0.112 0.2 45 tags AY AY ils ... 8952 079 0459 0828 61 79 =38

1 66 77 3 %, list=

21%, signa

1=47

%

KEGG JAK STAT SIGNA KEGG JAK STAT SIGNA Deta 10 0.27 1.08 0.28 0.329 0.8 23 tags LING PATHWAY LING PATHWAY ils ... 9 0877 604 2937 4853 61 64 =18

87 11 38 3 %, list=

11%, signa

1=20

%

Table 60: GSEA for BAF180-wildtype, BRGl-null vs. BAF180-wildtype, BRGl-wildtype

A704 cell lines on LL6 JAK STAT founder gene sets

3%, signal

=45%

HEMATOPOIETIN INTERFERON CLASSD200 DOM 25 0.332 0.932 0.5559 0.8389 1 470 tags=4 AIN CYTOKINE RECEPTOR ACTIVITY 1352 1065 9475 92 3 8%, list=2

1%, signal

=61%

GROWTH FACTOR BINDING 24 0.309 0.865 0.6749 0.8638 1 325 tags=2

0261 60124 3474 4636 7 9%, list=l

5%, signal

=34%

INTERLEUKIN BINDING 17 0.283 0.730 0.8388 0.8643 1 445 tags=4

4838 4276 889 178 9 1%, list=2

0%, signal

=51%

Table 6P: GSEA for BAF180-null vs. BAF180-wildtype A704 cell lines on E2F Founder gene sets

KAUFFMANN MELANOMA RELAPSE 60 0.390 1.408 0.058 0.257 0.999 6608 tags=48%, list=30%, UP 2316 2451 31533 6113 signal=69%

7 3

MODULE 54 25 0.324 1.465 0.004 0.259 0.987 4682 tags=29%, list=21%,

1 1876 012 93827 48274 signal=36%

7 2

PU J AN A XPRS S INT NETWORK 16 0.326 1.383 0.021 0.259 1 6911 tags=44%, list=31%,

5 9663 5881 02803 6924 signal=63%

2 8

REGULATION OF DNA REPLICATION 19 0.525 1.456 0.047 0.259 0.989 5881 tags=58%, list=26%,

6427 8967 19101 86636 signal=79% 5 3

BIOCARTA G 1 PATHWAY 27 0.463 1.409 0.085 0.260 0.998 2997 tags=26%, list=13%,

6605 431 53971 4318 signal=30%

7

MATZUK SPERMATOCYTE 68 0.396 1.467 0.023 0.261 0.984 3691 tags=24%, list=17%,

3679 985 64066 03795 signal=28% 7 2

SHEPARD BMYB MORPHOLINO DN 18 0.326 1.399 0.016 0.262 0.999 3851 tags=28%, list=17%,

0 0228 8913 86747 0695 signal=34%

BIOCARTA MCM PATHWAY 18 0.523 1.450 0.062 0.262 0.991 7686 tags=72%, list=35%,

641 6925 63982 82194 signal=110%

ISHIDA E2F TARGETS 51 0.405 1.410 0.046 0.264 0.998 6960 tags=59%, list=31%,

3819 2687 90831 1048 signal=85% 2 5

VANTVEER BREAST CANCER METAS 55 0.391 1.412 0.05 0.265 0.996 2904 tags=29%, list=13%, TASIS UP 082 3961 44183 signal=33%

ZHANG TLX TARGETS DN 88 0.368 1.414 0.023 0.267 0.996 5921 tags=43%, list=27%,

3990 3125 25581 21817 signal=59% 8 4

KEGG BASE EXCISION REPAIR 34 0.447 1.444 0.041 0.268 0.993 3979 tags=35%, list=18%,

5151 0593 48471 07487 signal=43%

7

KANG DOXORUBICIN RESISTANCE U 54 0.388 1.375 0.047 0.268 1 6814 tags=46%, list=31%, P 5366 6194 72727 1853 signal=67%

3 2

GNF2 RFC3 41 0.417 1.414 0.056 0.271 0.996 6900 tags=51%, list=31%,

5211 9994 89278 23934 signal=74% 8

REACTOME_G2_M_CHECKPOINTS 41 0.416 1.416 0.063 0.275 0.996 6732 tags=56%, list=30%,

7203 3488 45733 00415 signal=80%

SOTIRIOU BREAST CANCER GRADE 14 0.321 1.357 0.020 0.276 1 7618 tags=47%, list=34%, 1 VS 3 UP 9 4344 1204 78522 59324 signal=71%

ODONNELL TARGETS OF MYC AND 44 0.418 1.418 0.044 0.276 0.996 5684 tags=48%, list=26%, TFRC DN 8970 553 94382 91594 signal=64%

6

MARKEY RB l ACUTE LOF UP 22 0.303 1.358 0.014 0.278 1 5168 tags=31%, list=23%,

8 452 13 31980 82314 signal=40%

9

JOHANSSON GLIOMAGENESIS BY PD 55 0.382 1.350 0.074 0.278 1 2725 tags=18%, list=12%, GFB UP 1930 983 88987 9827 signal=21%

6

GNF2 BUB IB 49 0.392 1.348 0.095 0.279 1 6911 tags=43%, list=31%,

8091 7504 12761 3027 signal=62% 5

SIMBULAN PARP l TARGETS DN 17 0.521 1.420 0.074 0.279 0.996 4341 tags=47%, list=20%,

6608 628 23580 34334 signal=58%

4

GRAHAM CML QUIESCENT VS NOR 77 0.386 1.434 0.027 0.279 0.994 4926 tags=27%, list=22%, MAL QUIESCENT UP 1661 2515 58620 565 signal=35%

6 7

LI WILMS TUMOR 26 0.462 1.423 0.067 0.280 0.995 3332 tags=27%, list=15%,

0095 792 72009 35206 signal=32% 2

RIBONUCLEASE ACTIVITY 21 0.476 1.362 0.109 0.280 1 1098 tags=14%, list=5%,

3577 0135 27835 49374 signal=15% 3

REACTOME BASE EXCISION REPAIR 19 0.488 1.363 0.126 0.281 1 4761 tags=47%, list=21%,

5505 8277 88172 8866 signal=60% 4

STEIN ESRI TARGETS 80 0.355 1.358 0.065 0.281 1 4066 tags=30%, list=18%,

7025 726 02242 89817 signal=37% 2 4 GNF2_CCNA2 67 0.361 1.351 0.056 0.281 1 7686 tags=55%, list=35%,

1408 6709 76856 91626 signal=84% 8

ZHAN MULTIPLE MYELOMA PR UP 45 0.401 1.365 0.080 0.282 1 6377 tags=49%, list=29%,

3220 6914 56872 49145 signal=68% 7

SONG TARGETS OF IE86 CMV PROT 60 0.394 1.427 0.040 0.285 0.995 6608 tags=52%, list=30%, EIN 3113 9 86538 49019 signal=73%

7 4

BURTON ADIPOGENESIS PEAK AT 16 39 0.408 1.343 0.080 0.285 1 3108 tags=23%, list=14%, HR 9277 1141 17817 85017 signal=27%

7

GNF2 SMC2L1 32 0.447 1.423 0.052 0.286 0.995 6911 tags=50%, list=31%,

5485 9812 63158 8688 signal=72%

MODULE 403 45 0.394 1.337 0.080 0.291 1 4748 tags=36%, list=21%,

3862 5821 17817 4272 signal=45% 3

PYEON HPV POSITIVE TUMORS UP 88 0.341 1.334 0.047 0.292 1 3297 tags=27%, list=15%,

4747 982 05882 56615 signal=32% 7 4

WILCOX RESPONSE TO PROGESTERO 13 0.317 1.330 0.034 0.296 1 2958 tags=27%, list=13%,

NE UP 9 7961 7937 65346 98354 signal=31%

4 6

MANALO HYPOXIA DN 28 0.291 1.326 0.020 0.301 1 6851 tags=35%, list=31%,

4 3684 2402 88772 2804 signal=50%

8 9

BENPORATH PROLIFERATION 14 0.314 1.319 0.030 0.303 1 4658 tags=27%, list=21%,

4 8363 144 37974 10413 signal=34%

2 6

SGCGSSAAA_V$E2F1DP2_01 16 0.309 1.319 0.023 0.306 1 6280 tags=39%, list=28%,

2 7671 2339 41920 82164 signal=54%

6 3

NEGATIVE REGULATION OF CELL C 75 0.351 1.321 0.054 0.307 1 3835 tags=24%, list=17%, YCLE 5058 3702 8926 49267 signal=29%

5

BOYAULT LIVER CANCER SUBCLAS 44 0.380 1.297 0.102 0.310 1 1804 tags=16%, list=8%, S G123 UP 0448 2541 94118 85995 signal=17%

RB P130 DN.V1 UP 12 0.316 1.293 0.061 0.314 1 2543 tags=16%, list=ll%,

1 8952 6934 03286 2914 signal=18%

5

CHANG CYCLING GENES 14 0.311 1.297 0.069 0.314 1 6070 tags=41%, list=27%,

3 5874 2972 04762 42836 signal=56%

5

BAKER HEMATOPOIESIS STAT3 TAR 16 0.498 1.304 0.143 0.314 1 5606 tags=56%, list=25%, GETS 8392 5712 46895 92957 signal=75%

3

LY AGING OLD DN 54 0.372 1.308 0.084 0.315 1 3282 tags=22%, list=15%,

4437 7479 41558 1152 signal=26% 4 5

DNA METABOLIC PROCESS 24 0.288 1.306 0.018 0.315 1 4292 tags=25%, list=19%,

3 8049 2418 73536 527 signal=31%

5 2

OLSSON E2F3 TARGETS DN 44 0.392 1.310 0.108 0.315 1 2185 tags=16%, list=10%,

9537 525 84353 65085 signal=18% 5 5

DNA REPLICATION 98 0.326 1.299 0.067 0.316 1 6608 tags=45%, list=30%,

4416 9647 56756 94692 signal=64% 8 5

MODULE 485 49 0.374 1.297 0.108 0.317 1 5816 tags=39%, list=26%,

3328 536 59728 62144 signal=52%

6

GNF2 CKS IB 37 0.392 1.301 0.113 0.318 1 6911 tags=51%, list=31%,

1817 194 04348 87347 signal=74% 5

GARGALOVIC RESPONSE TO OXIDIZ 51 0.366 1.283 0.124 0.331 1 4223 tags=27%, list=19%, ED PHOSPHOLIPIDS TURQUOISE DN 1638 0538 71132 00462 signal=34%

5

CROONQUIST IL6 DEPRIVATION DN 97 0.318 1.276 0.071 0.332 1 6762 tags=43%, list=30%,

8555 713 07843 37627 signal=62% 5 5

V$E2F1_Q6_01 22 0.286 1.277 0.037 0.334 1 4825 tags=28%, list=22%,

9 8161 3455 03703 74997 signal=35%

7 AFFAR YY l TARGETS DN 21 0.291 1.277 0.045 0.338 3979 tags=25%, list=18%,

0 8410 4206 3408 ¹ signal=30%

6

MODULE 397 11 0.319 1.270 0.056 0.341 3317 tags=22%, list=15%,

1 4503 2408 87203 33938 ¹ signal=25%

2 6

NUCLEASE ACTIVITY 51 0.359 1.263 0.138 0.351 3059 tags=20%, list=14%,

8291 786 8889 47074 ¹ signal=23% 6

REACTOME ACTIVATION OF THE PR 30 0.401 1.259 0.173 0.357 6577 tags=57%, list=30%, E REPLICATIVE COMPLEX 0431 3498 91305 6477 ¹ signal=80%

2

GNF2 RRM2 40 0.376 1.253 0.129 0.362 7909 tags=58%, list=36%,

2977 236 93039 92648 ¹ signal=89% 7

KORKOLA TERATOMA 37 0.385 1.254 0.138 0.364 872 tag ll%,

4039 3806 8889 4212 ¹ s= list=4%, signal=ll%

VANTVEER BREAST CANCER POOR 51 0.359 1.250 0.128 0.365 5452 tags=43%, list=25%, PROGNOSIS 5011 7282 95928 02436 ¹ signal=57%

2

CROONQUIST NRAS SIGNALING DN 72 0.338 1.242 0.122 0.368 7393 tags=54%, list=33%,

0596 6128 49443 64632 ¹ signal=81% 3

RB P107 DN.V1 UP 13 0.298 1.244 0.087 0.368 3979 tags=28%, list=18%,

4 1448 0923 5 79972 ¹ signal=34%

8

REACTOME EXTENSION OF TELOME 27 0.409 1.244 0.176 0.371 3979 tags=33%, list=18%, RES 0032 3675 99115 91126 ¹ signal=41%

GROSS HYPOXIA VIA ELK3 ONLY D 44 0.360 1.244 0.150 0.374 3865 tags=27%, list=17%, N 8235 9167 34169 3939 ¹ signal=33%

4

ZHENG GLIOBLASTOMA PLASTICITY 23 0.274 1.228 0.050 0.377 4748 tags=28%, list=21%, UP 6 6216 8359 93833 14297 ¹ signal=35%

4 8

VERNELL RETINOBLASTOMA PATHW 70 0.328 1.237 0.122 0.377 6799 tags=44%, list=31%, AY UP 0144 3369 30216 2047 ¹ signal=64%

3

WONG EMBRYONIC STEM CELL CO 32 0.264 1.235 0.038 0.377 6597 tags=33%, list=30%, RE 7 8832 5665 35616 507 ¹ signal=47%

3

CELL CYCLE CHECKPOINT GO 00000 46 0.361 1.229 0.140 0.379 3946 tags=35%, list=18%, 75 7034 2565 96916 93726 ¹ signal=42%

BHATI G2M ARREST BY 2METHOXY 10 0.306 1.231 0.105 0.381 3735 tags=23%, list=17%, ESTRADIOL UP 7 0861 9682 38641 54486 ¹ signal=28%

5

AMUNDSON GENOTOXIC SIGNATURE 10 0.308 1.230 0.109 0.381 2516 tags=16%, list=ll%,

0 2465 115 1314 94412 ¹ signal=18%

8

RUIZ TNC TARGETS DN 13 0.292 1.222 0.072 0.383 5197 tags=35%, list=23%,

9 9560 6777 59953 599 ¹ signal=45%

8

PETROVA PROXl TARGETS UP 28 0.400 1.223 0.191 0.384 1909 tags=25%, list=9%,

8765 8188 30434 32854 ¹ signal=27%

MORI IMMATURE B LYMPHOCYTE D 88 0.315 1.215 0.130 0.388 4682 tags=28%, list=21%, N 0870 9486 13698 95854 ¹ signal=36%

8

PID RB 1PATHWAY 61 0.336 1.216 0.147 0.390 2997 tags=20%, list=13%,

7450 7755 7516 85585 ¹ signal=23% 2

KAUFFMANN DNA REPLICATION GE 13 0.292 1.217 0.109 0.393 4257 tags=21%, list=19%, NES 6 3885 2049 94764 48933 ¹ signal=26%

6 4

VECCHI GASTRIC CANCER EARLY U 40 0.251 1.202 0.036 0.417 2649 tags=13%, list=12%, P 3 9435 6228 74540 13816 ¹ signal=15%

6 7

BIOCARTA P53 PATHWAY 16 0.445 1.189 0.257 0.419 9112 tags=69%, list=41%,

2219 8873 38397 01195 ¹ signal=116%

NEGATIVE REGULATION OF DNA M 17 0.452 1.195 0.231 0.420 2026 tags=24%, list=9%, ETABOLIC PROCESS 444 021 78808 93435 ¹ signal=26%

V$E2F 01 63 0.320 1.191 0.176 0.421 5589 tags=33%, list=25%,

2585 4719 87075 99737 ¹ signal=44%

3 OXFORD RALA OR RALB TARGETS 48 0.352 1.196 0.195 0.422 6184 tags=44%, list=28%,

UP 8961 0196 27897 1113 ¹ signal=60%

5

MODULE 325 51 0.335 1.192 0.193 0.422 4349 tags=27%, list=20%,

2738 8161 76393 3233 ¹ signal=34% 6

YAO TEMPORAL RESPONSE TO PRO 30 0.379 1.189 0.195 0.422 1896 tags=13%, list=9%,

GESTERONE CLUSTER l 5 5156 9841 02075 3635 ¹ signal=15%

5

BENPORATH ES CORE NINE CORREL 94 0.303 1.196 0.139 0.425 4666 tags=29%, list=21%, ATED 5604 2134 90825 2052 ¹ signal=36%

7

V$E2F1_Q4 23 0.270 1.197 0.063 0.425 3607 tags=18%, list=16%,

5 0731 4943 88206 86854 ¹ signal=22%

2

MODULE 252 23 0.263 1.183 0.082 0.431 6802 tags=39%, list=31%,

5 6046 151 32445 19183 ¹ signal=56%

4

GNF2 ESPL1 35 0.355 1.179 0.211 0.433 6911 tags=51%, list=31%,

1145 1232 49425 50247 ¹ signal=75% 8

MODULE 57 54 0.338 1.180 0.192 0.434 4926 tags=26%, list=22%,

5203 2068 22462 5493 ¹ signal=33%

CELL_CYCLE_ARREST_GO_0007050 53 0.325 1.163 0.211 0.446 3835 tags=19%, list=17%,

9060 6739 98156 74337 ¹ signal=23% 7

WINNEPENNINCKX MELANOMA MET 16 0.275 1.164 0.127 0.448 7430 tags=44%, list=33%, ASTASISJJP 0 8405 3463 87724 5678 ¹ signal=65%

2

GCNP SHH UP LATE.VI UP 17 0.269 1.165 0.145 0.449 4748 tags=28%, list=21%,

3 6331 314 63107 76816 ¹ signal=35%

4

V$E2F1_Q4_01 21 0.263 1.170 0.109 0.449 5320 tags=30%, list=24%,

9 1721 919 72568 92062 ¹ signal=39%

2 4

WHITFIELD CELL CYCLE GI S 13 0.281 1.167 0.166 0.452 3471 tags=22%, list=16%,

4 6848 068 27635 56835 ¹ signal=25%

2

REGULATION OF MITOTIC CELL CY 23 0.399 1.165 0.232 0.452 4815 tags=39%, list=22%,

CLE 2518 4276 84823 96666 ¹ signal=50%

8

WANG RESPONSE TO GSK3 INHIBIT 34 0.251 1.167 0.101 0.454 5102 tags=30%, list=23%, OR SB216763 DN 5 9045 4849 06383 85333 ¹ signal=38%

2

V$E2F_Q3 21 0.262 1.158 0.142 0.457 5144 tags=28%, list=23%,

2 1424 2417 1801 13925 ¹ signal=36%

V$E2F_Q6_01 22 0.257 1.152 0.122 0.462 3595 tags=21%, list=16%,

7 5921 2388 00957 74084 ¹ signal=24%

SARRIO EPITHELIAL MESENCHYMAL 16 0.265 1.153 0.153 0.463 4703 tags=31%, list=21%, TRANSITION UP 9 0023 0817 84616 9936 ¹ signal=39%

7

DNA POLYMERASE ACTIVITY 17 0.422 1.153 0.276 0.466 722 tags=12%, list=3%,

6581 4702 68846 5431 ¹ signal=12%

GRAHAM CML DIVIDING VS NORMA 16 0.267 1.149 0.187 0.466 7422 tags=45%, list=33%, L QUIESCENT UP 4 7485 2176 34793 72526 ¹ signal=67%

PID BARD 1 PATHWAY 29 0.370 1.146 0.264 0.469 3835 tags=24%, list=17%,

7223 7375 06926 454 ¹ signal=29% 8

CHIANG LIVER CANCER SUBCLASS 16 0.266 1.144 0.169 0.471 5315 tags=28%, list=24%, PROLIFERATION UP 8 8343 6294 45107 34057 ¹ signal=36%

8

PID FOXM1 PATHWAY 39 0.344 1.137 0.252 0.476 4586 tags=28%, list=21%,

2356 4557 25225 67563 ¹ signal=35%

LI WILMS TUMOR VS FETAL KIDNE 16 0.269 1.137 0.184 0.479 3282 tags=14%, list=15%, Y 1 DN 1 0813 7878 51025 2994 ¹ signal=16%

8

WHITFIELD CELL CYCLE G2 17 0.263 1.134 0.158 0.480 7199 tags=38%, list=32%,

3 1934 8187 01887 18038 ¹ signal=55%

GARCIA TARGETS OF FLU AND DA 16 0.265 1.138 0.182 0.482 6630 tags=38%, list=30%, XI DN 4 1164 0086 66979 10818 ¹ signal=54%

2

REGULATION OF DNA METABOLIC P 43 0.342 1.132 0.253 0.483 4815 tags=37%, list=22%, ROCESS 7879 1517 48836 24963 ¹ signal=47%

5 LEE EARLY T LYMPHOCYTE UP 97 0.250 0.986 0.472 0.716 5139 tags=32%, list=23%,

1194 5701 5537 96126 ¹ signal=41% 5

KAUFFMANN DNA REPAIR GENES 21 0.222 0.984 0.522 0.718 3979 tags=20%, list=18%,

9 108 6902 5653 36126 ¹ signal=24%

SHEPARD CRUSH AND BURN MUTA 16 0.229 0.980 0.509 0.723 4748 tags=27%, list=21%, NT DN 4 1851 4757 0909 80483 ¹ signal=34%

5

V$E2F_Q4_01 22 0.220 0.981 0.501 0.724 5320 tags=30%, list=24%,

7 2432 5795 17093 19137 ¹ signal=38%

3

REACTOME FANCONI ANEMIA PATH 21 0.338 0.977 0.490 0.730 8209 tags=52%, list=37%, WAY 1209 3252 82568 0886 ¹ signal=83%

4 6

HOFFMANN LARGE TO SMALL PRE 15 0.229 0.966 0.523 0.748 6690 tags=39%, list=30%, BII LYMPHOCYTE UP 5 3058 305 918 9484 ¹ signal=55%

PID FANCONI PATHWAY 47 0.285 0.966 0.507 0.750 6732 tags=47%, list=30%,

0136 9817 5269 541 ¹ signal=67% 2 7

PIONTEK PKD 1 TARGETS DN 16 0.361 0.963 0.506 0.751 3510 tags=38%, list=16%,

1944 1549 383 2874 ¹ signal=45% 6

CHROMATIN 33 0.300 0.963 0.505 0.753 2919 tags=18%, list=13%,

6569 6298 3996 42274 ¹ signal=21% 4

FERREIRA EWINGS SARCOMA UNST 15 0.225 0.967 0.545 0.753 6650 tags=36%, list=30%, ABLE VS STABLE UP 9 9697 0589 2323 96377 ¹ signal=51%

3

GROSS HYPOXIA VIA ELK3 UP 20 0.219 0.968 0.558 0.754 3787 tags=19%, list=17%,

4 5860 2438 753 3387 ¹ signal=23%

6 4

SCIBETTA_KDM5B_TARGETS_DN 77 0.252 0.958 0.551 0.761 5602 tags=32%, list=25%,

8328 3655 33927 84374 ¹ signal=43% 3

EGUCHI CELL CYCLE RB 1 TARGETS 23 0.322 0.948 0.527 0.773 7686 tags=57%, list=35%,

7037 5936 64976 58365 ¹ signal=86% 8

WHITFIELD_CELL_CYCLE_G2_M 21 0.213 0.949 0.627 0.774 3989 tags=20%, list=18%,

1 3046 3208 8481 9125 ¹ signal=24%

5

RPS14 DN.V1 DN 17 0.218 0.950 0.589 0.776 4815 tags=28%, list=22%,

8 9350 1130 1648 29346 ¹ signal=35%

7 6

MOLENAAR TARGETS OF CCND1 AN 57 0.267 0.952 0.527 0.777 7382 tags=49%, list=33%, D CDK4 DN 3039 2256 1493 40884 ¹ signal=73%

V$E2F1_Q6 22 0.212 0.950 0.596 0.779 3544 tags=19%, list=16%,

5 8831 3851 53467 0583 ¹ signal=23%

4 5

BOYAULT LIVER CANCER SUBCLAS 18 0.215 0.939 0.629 0.795 6309 tags=31%, list=28%, S_G3_UP 7 3137 9698 6296 1581 ¹ signal=43%

9 4

RHODES UNDIFFERENTIATED CANCE 68 0.249 0.936 0.589 0.801 7422 tags=43%, list=33%, R 8641 3587 2473 9492 ¹ signal=64%

9

VANTVEER BREAST CANCER METAS 11 0.228 0.933 0.608 0.806 4748 tags=25%, list=21%, TASIS DN 6 7171 5559 6956 67114 ¹ signal=32%

2

GNF2 CDC2 61 0.254 0.931 0.561 0.808 7559 tags=44%, list=34%,

2479 5171 9048 7767 ¹ signal=67% 3

ODONNELL TFRC TARGETS DN 11 0.228 0.928 0.629 0.809 1958 tags=12%, list=9%,

7 9518 8347 54545 48067 ¹ signal=13%

7

GNF2 CCNB2 56 0.262 0.929 0.603 0.810 7559 tags=45%, list=34%,

4212 6985 93876 3493 ¹ signal=67% 5

MODULE 123 22 0.207 0.926 0.685 0.813 3622 tags=19%, list=16%,

5 2971 3362 50366 29805 ¹ signal=23%

MITSIADES RESPONSE TO APLIDIN 24 0.206 0.915 0.721 0.825 3967 tags=18%, list=18%, DN 4 6100 9852 8045 3738 ¹ signal=21%

7

BIOCARTA CELLCYCLE PATHWAY 22 0.314 0.913 0.584 0.827 3518 tags=36%, list=16%,

4141 8003 74576 61246 ¹ signal=43%

6 PETROVA ENDOTHELIUM LYMPHATI 12 0.224 0.916 0.681 0.827 4748 tags=30%, list=21%, C VS BLOOD UP 5 5195 3942 0551 8902 ¹ signal=37%

4

BIOCARTA G2 PATHWAY 24 0.308 0.918 0.555 0.828 4926 tags=25%, list=22%,

0662 9605 81397 4411 ¹ signal=32% 8

CHANG CORE SERUM RESPONSE UP 20 0.207 0.911 0.713 0.829 3331 tags=16%, list=15%,

5 9613 8108 9423 64015 ¹ signal=19%

5

WEST ADRENOCORTICAL TUMOR UP 28 0.199 0.909 0.762 0.831 3011 tag l%, list=14%,

8 5358 7519 2739 3348 ¹ s=l

signal=13%

3 5

KONG E2F3 TARGETS 93 0.233 0.916 0.629 0.831 7618 tags=53%, list=34%,

3433 4501 18663 5079 ¹ signal=80%

4

CONDENSED NUCLEAR CHROMOSO 18 0.323 0.908 0.597 0.831 541 tags ll%, list=2%,

ME 6068 3769 3742 60794 ¹ =

signal=ll%

3

REACTOME E2F MEDIATED REGULA 31 0.286 0.919 0.582 0.831 3225 tags=23%, list=15%,

TION OF DNA REPLICATION 8581 1460 2222 6486 ¹ signal=26%

7 6

NAKAMURA CANCER MICROENVIRO 45 0.261 0.901 0.625 0.835 4815 tags=24%, list=22%, NMENT_DN 4644 4584 5533 ¹ signal=31%

2 4

DNA INTEGRITY CHECKPOINT 22 0.312 0.902 0.603 0.837 2295 tags=23%, list=10%,

0776 1031 8136 4959 ¹ signal=25% 4

LE EGR2 TARGETS UP 10 0.225 0.903 0.683 0.838 7145 tags=39%, list=32%,

7 6782 4158 8565 00644 ¹ signal=58%

8

GOLDRATH ANTIGEN RESPONSE 31 0.196 0.904 0.788 0.839 2823 tags=14%, list=13%,

5 1421 1977 1356 6098 ¹ signal=16%

5

CELL CYCLE PROCESS 18 0.207 0.894 0.740 0.850 3835 tags=18%, list=17%,

1 3305 3612 099 8811 ¹ signal=21%

7 6

GEORGES CELL CYCLE MIR192 TAR 61 0.243 0.884 0.678 0.855 5137 tags=34%, list=23%, GETS 9715 5039 5714 3726 ¹ signal=45%

7 6

MODULE 337 59 0.246 0.890 0.677 0.856 4553 tags=25%, list=20%,

3871 7535 48916 8585 ¹ signal=32% 2

CHROMOSOME 11 0.216 0.884 0.726 0.858 3952 tags=19%, list=18%,

9 2617 5944 16136 7043 ¹ signal=23%

7

CSR LATE UP.V1 UP 16 0.208 0.887 0.759 0.860 6013 tags=35%, list=27%,

2 8422 9334 7254 4634 ¹ signal=47%

6 3

DNA REPAIR 12 0.217 0.886 0.761 0.860 4292 tags=21%, list=19%,

1 6508 4344 9048 91167 ¹ signal=26%

9 4

NUCLEAR CHROMOSOME 52 0.249 0.884 0.656 0.861 3753 tags=23%, list=17%,

8287 9984 4417 2027 ¹ signal=28% 4

NADERI BREAST CANCER PROGNOSI 45 0.251 0.878 0.690 0.867 3941 tags=24%, list=18%, S_UP 694 2624 3226 421 ¹ signal=30%

6

WEST ADRENOCORTICAL TUMOR M 20 0.310 0.872 0.644 0.877 4586 tags=35%, list=21%,

ARKERS UP 1387 6998 2953 3783 ¹ signal=44%

6

DNA RECOMBINATION 41 0.261 0.864 0.696 0.882 3979 tags=24%, list=18%,

4301 5332 9697 53117 ¹ signal=30%

7

MODULE 98 38 0.183 0.864 0.910 0.885 6383 tags=30%, list=29%,

3 8496 5767 4859 9621 ¹ signal=41%

5

RESPONSE TO DNA DAMAGE STIMU 15 0.205 0.866 0.788 0.888 4292 tags=21%, list=19%, LUS 5 1644 7723 8349 27926 ¹ signal=25%

2 5

PID AURORA B PATHWAY 38 0.259 0.865 0.689 0.888 7282 tags=39%, list=33%,

2319 0617 8148 4687 ¹ signal=59% 8 6 PID ATR PATHWAY 38 0.176 0.575 0.985 0.997 7393 tags=42%, list=33%,

2004 4561 9719 92004 ¹ signal=63%

4

XU HGF TARGETS INDUCED BY AK 23 0.208 0.604 0.961 0.998 6431 tags=26%, list=29%, Tl 48HR DN 75 0176 96866 7243 ¹ signal=37%

PUJANA BREAST CANCER LIT INT N 10 0.145 0.586 1 0.999 4926 tags=20%, list=22%, ETWORK 0 6802 1298 418 ¹ signal=26%

6

PID AURORA A PATHWAY 31 0.130 0.405 1 0.999 7863 tags=35%, list=35%,

3978 1939 6472 ¹ signal=55% 9 5

REACTOME DOUBLE STRAND BREA 22 0.217 0.633 0.925 0.999 3979 tags=27%, list=18%, K REPAIR 2065 8607 9259 9909 ¹ signal=33%

4 7

BOYAULT LIVER CANCER SUBCLAS 52 0.210 0.732 0.905 5137 tags=35%, list=23%, S G23 UP 1307 1246 31176 ^{1 1} signal=45%

8

WAKASUGI HAVE ZNF143 BINDING 57 0.200 0.724 0.932 2548 t

SITES 7666 2195 1267 ^{1 1} ags=ll%, list=ll%, signal=12% 6

STRUCTURE SPECIFIC DNA BINDING 55 0.201 0.719 0.912 4340 tags=24%, list=20%,

8936 3361 6214 ^{1 1} signal=29% 9

RNA CATABOLIC PROCESS 21 0.249 0.715 0.857 1091 tags=10%, list=5%,

1310 3406 4514 ^{1 1} signal=10% 2

JUBAN TARGETS OF SPI1 AND FLU 85 0.185 0.709 0.951 3630 tags=18%, list=16%, DN 801 6551 94507 ^{1 1} signal=21%

7

M PHASE 10 0.177 0.699 0.967 3835 tags=18%, list=17%,

4 3396 1719 8161 ^{1 1} signal=22%

9

SLEBOS HEAD AND NECK CANCER 79 0.181 0.697 0.967 6611 tags=37%, list=30%, WITH HPV UP 0667 8613 8161 ^{1 1} signal=52%

7

LE NEURONAL DIFFERENTIATION D 19 0.243 0.683 0.873 2707 tags=16%, list=12%,

N 1075 0966 50833 ^{1 1} signal=18%

3

MODULE 244 18 0.156 0.676 1 4989 tags=18%, list=22%,

3 1229 9737 ^{1 1} signal=23%

8

KEGG HOMOLOGOUS RECOMBINATI 26 0.218 0.669 0.920 3835 tags=27%, list=17%, ON 1942 2233 50207 ^{1 1} signal=33%

3

DEOXYRIBONUCLEASE ACTIVITY 22 0.226 0.663 0.931 3059 tags=18%, list=14%,

4191 2232 1111 ^{1 1} signal=21%

MITOTIC CELL CYCLE 14 0.155 0.659 0.997 4815 tags=19%, list=22%,

9 7022 1574 7477 ^{1 1} signal=24%

5

CONDENSED CHROMOSOME 33 0.206 0.657 0.938 541 tags=6%, list=2%,

7506 9867 0734 ^{1 1} signal=6% 5

LY AGING MIDDLE DN 16 0.246 0.649 0.882 5139 tags=31%, list=23%,

9101 7035 35295 ^{1 1} signal=41%

7 6

CHROMATIN BINDING 30 0.206 0.647 0.947 6704 tags=40%, list=30%,

2900 8348 2477 ^{1 1} signal=57% 4 4

FINETTI BREAST CANCER KINOME 16 0.236 0.644 0.893 8991 tags=63%, list=40%, RED 1942 3034 0818 ^{1 1} signal=105%

2 4

NEMETH INFLAMMATORY RESPONS 30 0.202 0.636 0.964 1267 tags=7%, list=6%, E LPS DN 7904 2227 44446 ^{1 1} signal=7%

5

MODULE 372 23 0.211 0.623 0.953 2603 tags=17%, list=12%,

8640 1687 53985 ^{1 1} signal=20% 8 7

LI WILMS TUMOR ANAPLASTIC UP 18 0.178 0.480 0.989 4586 tags=22%, list=21%,

6921 8827 40676 ^{1 1} signal=28% Table 6Q: GSEA for BAF180-wildtype, BRGl-null vs. BAF180-wildtype, BRGl-wildtype

A704 cell lines on E2F founder gene sets

REACTOME DNA STRAND ELONGATIO 30 0.466 1.365 0.0917431 0.446354 1 2108 tags=20%,

N 1894 0514 2 57 list=9%,

4 signal=22%

V$E2F_Q3_01 22 0.326 1.242 0.0701754 0.446707 1 5052 tags=30%,

6 2826 3823 4 55 list=23%, 5 signal=38%

CELL_CYCLE_CHECKPOINT_GO_0000075 47 0.444 1.419 0.0410132 0.447200 1 6550 tags=47%,

5298 703 7 95 list=29%,

6 signal=66%

CELL CYCLE GO 0007049 30 0.332 1.301 0.0223577 0.447624 1 6550 tags=35%,

0 9150 0465 23 4 list=29%, 4 signal=49%

REACTOME HOMOLOGOUS RECOMBIN 16 0.486 1.242 0.1862068 0.449743 1 6871 tags=56%,

ATION REPAIR OF REPLICATION INDE 2423 8929 9 4 list=31%,

PENDENT DOUBLE STRAND BREAKS 5 signal=81%

MODULE 403 46 0.415 1.302 0.1169154 0.450989 1 7716 tags=54%,

0462 2362 2 1 list=35%,

4 signal=83%

ODONNELL TARGETS OF MYC AND TF 44 0.414 1.285 0.1175757 0.451063 1 8499 tags=59%,

RC DN 673 5059 6 6 list=38%, signal=96%

WHITFIELD CELL CYCLE G2 17 0.327 1.227 0.0887011 0.451402 1 4668 tags=24%,

3 7359 4308 6 list=21%, 3 signal=30%

DNA DAMAGE CHECKPOINT 19 0.474 1.243 0.2043010 0.452576 1 6550 tags=58%,

7511 4118 7 9 list=29%,

7 signal=82%

CHIANG LIVER CANCER SUBCLASS PR 16 0.331 1.244 0.0827147 0.452887 1 4651 tags=24%,

OLIFERATION UP 8 6217 898 44 95 list=21%,

4 signal=30%

MARKEY RB 1 CHRONIC LOF UP 10 0.351 1.245 0.1239035 0.455226 1 4537 tags=30%,

7 6871 7684 06 list=20%, signal=37%

REACTOME_G2_M_CHECKPOINTS 41 0.428 1.309 0.1253196 0.456026 1 9054 tags=66%,

0654 6297 9 32 list=41%, signal=lll%

PID FANCONI PATHWAY 47 0.446 1.422 0.0501835 0.456052 1 8291 tags=64%,

8509 3179 98 84 list=37%,

6 signal=102%

GNF2 SMC4L1 84 0.359 1.246 0.1190751 0.456960 1 8761 tags=46%,

6912 8225 4 02 list=39%, signal=76%

V$E2F1_Q3 23 0.336 1.285 0.0413223 0.457054 1 6428 tags=35%,

1 029 5971 13 5 list=29%, signal=49%

V$E2F1DP1RB_01 22 0.341 1.302 0.0269430 0.457450 1 5052 tags=30%,

0 1136 4278 06 27 list=23%, signal=38%

RESPONSE TO DNA DAMAGE STIMULU 15 0.350 1.304 0.0465872 0.458742 1 8304 tags=50%,

S 6 5024 2861 14 86 list=37%,

3 signal=79%

V$E2F1_Q6_01 23 0.370 1.411 0.0104058 0.459113 1 7781 tags=46%,

0 6371 9506 27 48 list=35%, 8 signal=70%

BLUM RESPONSE TO SALIRASIB DN 33 0.333 1.310 0.0132248 0.459269 1 4415 tags=25%,

3 1813 9615 22 88 list=20%, signal=30%

CHROMATIN 33 0.428 1.290 0.1581769 0.460228 1 4588 tags=39%,

6325 6651 4 83 list=21%,

3 signal=50%

V$E2F1_Q4 23 0.337 1.292 0.0340206 0.461104 1 4570 tags=25%,

2 286 5217 18 48 list=21%, signal=32%

V$E2F1DP1_01 22 0.333 1.286 0.0351602 0.461503 1 4539 tags=27%,

7 0221 2234 9 12 list=20%, signal=33%

DNA POLYMERASE ACTIVITY 17 0.511 1.305 0.1591220 0.463127 3049 tags=29%,

2989 1121 8 02 ¹ list=14%, 5 signal=34%

REACTOME ACTIVATION OF THE PRE 30 0.421 1.221 0.1932555 0.463424 1 9054 tags=60%,

REPLICATIVE COMPLEX 6452 9079 1 62 list=41%,

8 signal=101% LINDGREN BLADDER CANCER CLUSTE 35 0.290 1.138 0.1523713 0.588673 1 5957 tags=30%,

R 1 DN 9 4739 0037 5 5 list=27%, signal=41%

V$E2F_01 65 0.334 1.111 0.2972973 0.596143 6939 tags=38%,

9032 9276 9 ¹ list=31%, 7 signal=56%

GEORGES CELL CYCLE MIR192 TARGE 61 0.338 1.112 0.3084223 0.598916 1 8144 tags=52%,

TS 8449 2179 4 list=37%,

8 signal=83%

GCNP_SHH_UP_LATE.V1_UP 17 0.300 1.133 0.2396166 0.599856 1 6242 tags=33%,

1 7716 2275 2 2 list=28%, 8 signal=45%

PID FOXMl PATHWAY 39 0.368 1.125 0.3072916 0.600943 1 5440 tags=36%,

5086 2751 6 4 list=24%,

4 signal=47%

M PHASE 10 0.312 1.114 0.2886363 0.601072 1 6500 tags=31%,

7 0373 0449 6 list=29%, 5 signal=43%

KTGGYRSGAA UNKNOWN 73 0.326 1.112 0.295612 0.602530 7550 tags=45%,

3472 2824 54 ¹ list=34%, 6 signal=68%

LY AGING PREMATURE DN 29 0.388 1.127 0.3226632 0.602791 1 2915 tags=17%,

2155 2681 5 67 list=13%, signal=20%

CHANG CYCLING GENES 14 0.308 1.131 0.2454252 0.602923 1 4768 tags=24%,

3 1746 0683 9 list=21%,

7 signal=31%

RIBONUCLEASE ACTIVITY 22 0.408 1.125 0.3163686 0.603118 1 6343 tags=32%,

1125 8738 4 6 list=29%,

3 signal=44%

REGULATION OF DNA METABOLIC PR 43 0.357 1.114 0.3061480 0.603185 1 6705 tags=44%,

OCESS 4490 5996 5 65 list=30%,

8 signal=63%

MORI PRE BI LYMPHOCYTE UP 76 0.327 1.129 0.2862069 0.603563 4570 tags=22%,

4124 6207 25 ¹ list=21%, signal=28%

DNA RECOMBINATION 41 0.362 1.115 0.296343 0.604119 7474 tags=56%,

7869 4684 2 ¹ list=34%, 2 signal=84%

WANG CISPLATIN RESPONSE AND XPC 18 0.297 1.127 0.2330198 0.606506 1 5313 tags=29%,

UP 4 9630 3884 6 17 list=24%,

5 signal=38%

GNF2 PCNA 67 0.329 1.115 0.3015494 0.606930 1 9683 tags=49%,

3745 769 6 85 list=44%, 8 signal=87%

STEIN ESRI TARGETS 81 0.317 1.106 0.3129411 0.607865 6467 tags=40%,

7997 642 6 8 ¹ list=29%, 8 signal=56%

REACTOME DOUBLE STRAND BREAK 22 0.403 1.104 0.3378196 0.608024 1 8192 tags=59%,

REPAIR 9069 1609 4 2 list=37%,

4 signal=94%

RB P107 DN.V1 UP 13 0.301 1.102 0.2899022 0.608211 1 4879 tags=29%,

3 7946 9329 7 7 list=22%, 2 signal=37%

FINETTI BREAST CANCER KINOME RE 16 0.444 1.121 0.3478261 0.609909 1 7061 tags=50%,

D 0954 1063 53 list=32%, signal=73%

MITSIADES RESPONSE TO APLIDIN DN 24 0.291 1.118 0.2371134 0.609931 1 7790 tags=37%,

3 6716 732 1 list=35%, 6 signal=56%

ROSTY CERVICAL CANCER PROLIFERA 13 0.300 1.116 0.2541436 0.609947 1 6224 tags=29%,

TION CLUSTER 9 7891 0735 6 26 list=28%,

2 signal=41%

JOHANSSON GLIOMAGENESIS BY PDGF 55 0.340 1.104 0.3160434 0.611175 1 7070 tags=42%,

B UP 3785 3452 4 3 list=32%,

2 signal=61%

RESPONSE TO ENDOGENOUS STIMULU 19 0.294 1.116 0.2277019 0.611684 1 6871 tags=38%, s 0 8509 8075 9 56 list=31%,

8 signal=54%

GNF2 FEN1 56 0.343 1.119 0.3089820 0.612732 1 7790 tags=36%,

0392 0727 4 77 list=35%,

4 signal=55% MISSIAGLIA REGULATED BY METHYL 11 0.253 0.917 0.6426193 0.898400 8599 tags=40%, ATION DN 7 8870 50246 7 ¹ list=39%,

3 signal=65%

PID E2F PATHWAY 72 0.265 0.892 0.6639248 0.900871 7061 tags=36%,

1171 4725 4 ¹ list=32%,

7 signal=53%

GNF2 RRM1 87 0.262 0.917 0.6188341 0.901395 1012 tags=47%,

0543 829 1 ¹ 4 list=46%, 5 signal=86%

DORMOY ELAVL 1 TARGETS 16 0.352 0.899 0.6144244 0.902217 4377 tags=31%,

1545 95354 1 ¹ list=20%, 5 signal=39%

CHROMOSOME 11 0.247 0.894 0.6868132 0.904068 7750 tags=37%,

9 2632 2441 2 ¹ list=35%, 7 signal=56%

E2F1 UP.V1 UP 18 0.236 0.892 0.7221052 0.904473 8039 tags=34%,

2 1003 5108 6 8 ¹ list=36%, 2 signal=52%

REACTOME PROCESSIVE SYNTHESIS 0 15 0.354 0.897 0.6262203 0.904643 4765 tags=27%, N THE LAGGING STRAND 5375 30215 5 2 ¹ list=21%,

8 signal=34%

RNA CATABOLIC PROCESS 21 0.328 0.895 0.6258503 0.904870 6343 tags=38%,

2475 5597 8 ¹ list=29%, 8 signal=53%

KEGG MISMATCH REPAIR 23 0.316 0.883 0.6328947 0.905280 2063 tags=13%,

8685 9419 5 65 ¹ list=9%,

7 signal=14%

SASAKI ADULT T CELL LEUKEMIA 16 0.238 0.888 0.7296137 0.907170 6939 tags=32%,

8 2873 1789 53 ¹ list=31%, 6 signal=46%

GNF2 BUB IB 49 0.275 0.886 0.6834532 0.907627 7790 tags=33%,

2565 18875 6 46 ¹ list=35%, signal=50%

NUCLEAR CHROMOSOME 52 0.275 0.884 0.6582430 0.907974 7686 tags=40%,

9821 33754 6 4 ¹ list=35%, 4 signal=62%

YU MYC TARGETS UP 42 0.282 0.880 0.6594663 0.909507 7442 tags=38%,

6596 31185 04 ¹ list=33%, 5 signal=57%

NAKAMURA CANCER MICROENVIRON 45 0.279 0.877 0.6662591 0.911936 4904 tags=18%, MENT DN 6858 47896 7 76 ¹ list=22%,

5 signal=23%

MITOSIS 80 0.253 0.872 0.7086705 0.918922 3765 tags=16%,

0464 46853 4 ¹ list=17%, signal=19%

PAL PRMT5 TARGETS UP 20 0.228 0.863 0.7713987 0.918950 7104 tags=31%,

0 4962 8039 7 ¹ list=32%, 4 signal=45%

LY AGING OLD DN 55 0.264 0.861 0.7130647 0.920155 6195 tags=25%,

9324 5848 3 35 ¹ list=28%, 2 signal=35%

DNA REPLICATIONJNITIATION 16 0.339 0.869 0.6421499 0.920635 9220 tags=69%,

6224 7947 04 ¹ list=42%, 7 signal=117%

LI WILMS TUMOR VS FETAL KIDNEY 29 0.295 0.863 0.6778523 0.922355 4765 tags=24%, 2_UP 3956 9146 3 3 ¹ list=21%,

4 signal=31%

MODULE 54 25 0.224 0.865 0.7917098 0.922756 4765 tags=20%,

0 5443 4779 4 1 ¹ list=21%, 9 signal=25%

KANG DOXORUBICIN RESISTANCE UP 54 0.267 0.867 0.6890756 0.922868 8499 tags=37%,

2452 11955 5 7 ¹ list=38%, 6 signal=60%

M PHASE OF MITOTIC CELL CYCLE 83 0.244 0.854 0.7244546 0.925178 3765 tags=16%,

7429 0637 4 ¹ list=17%, 6 signal=19%

FARMER_BREAST_CANCER_CLUSTER_2 33 0.286 0.850 0.6847682 0.925446 6747 tags=30%,

4472 2835 87 ¹ list=30%, 6 signal=43%

CONCANNON APOPTOSIS BY EPOXOMI 15 0.230 0.852 0.780065 0.925647 3636 tags=19%, CIN DN 5 4101 01883 1 ¹ list=16%, signal=22% XU HGF SIGNALING NOT VIA AKTl 48 20 0.228 0.618 0.9233871 4372 tags=20%, HR DN 7609 9885 ^{1 1} list=20%, signal=25%

MODULE 320 20 0.227 0.616 0.9306667 1052 tags=65%,

4218 54776 ^{1 1} 6 list=47%, 8 signal=123%

GNF2 CDC2 61 0.179 0.593 0.9847775 9937 tags=39%,

5068 2809 ^{1 1} list=45%, 8 signal=71%

WEST ADRENOCORTICAL TUMOR MAR 20 0.214 0.576 0.9659400 7442 tags=40%,

KERS UP 8801 8052 6 ^{1 1} list=33%,

2 signal=60%

REPLICATION FORK 18 0.217 0.571 0.9441417 8291 tags=44%,

0146 30104 ^{1 1} list=37%,

7 signal=71%

EGUCHI CELL CYCLE RB 1 TARGETS 23 0.204 0.564 0.9628647 7686 tags=30%,

7849 02063 ^{1 1} list=35%, 7 signal=46%

MORI LARGE PRE BII LYMPHOCYTE U 84 0.160 0.555 1 9054 tags=38%, P 8009 53854 ^{1 1} list=41%,

8 signal=64%

GNF2 ESPL1 35 0.185 0.553 0.9898089 1119 tags=51%,

8090 832 ^{1 1} 8 list=50%, 3 signal=104%

GREENBAUM_E2A_TARGETS_UP 33 0.182 0.551 0.9827814 6069 tags=24%,

3956 1335 7 ^{1 1} list=27%, 8 signal=33%

CONDENSED CHROMOSOME 33 0.184 0.548 0.9789082 667 tags=6%,

6433 1924 ^{1 1} list=3%, 1 signal=6%

GNF2 CENPE 40 0.178 0.545 0.981203 1002 tags=40%,

6746 7817 ^{1 1} 5 list=45%, 8 signal=73%

SMID BREAST CANCER LUMINAL A D 16 0.211 0.545 0.9726402 6195 tags=19%, N 7111 03566 ^{1 1} list=28%,

8 signal=26%

GNF2 CKS IB 37 0.175 0.539 0.9898348 8422 tags=30%,

3625 01255 5 ^{1 1} list=38%, 4 signal=48%

PENG GLUCOSE DEPRIVATION DN 16 0.138 0.520 1 8176 tags=29%,

0 5064 1427 ^{1 1} list=37%,

7 signal=46%

GNF2 MKI67 27 0.180 0.517 0.9797023 9937 tags=44%,

8232 92157 ^{1 1} list=45%, 4 signal=80%

CHROMOSOME SEGREGATION 32 0.172 0.515 0.9884020 9683 tags=44%,

4604 44017 7 ^{1 1} list=44%,

7 signal=77%

FRASOR RESPONSE TO SERM OR FUL 50 0.159 0.512 0.9951100 8003 tags=26%, VESTRANT DN 1275 43126 3 ^{1 1} list=36%,

8 signal=41%

GNF2 H2AFX 31 0.162 0.480 0.9974716 3247 tags=10%,

9448 6161 ^{1 1} list=15%, 2 signal=ll%

GNF2 CCNB2 56 0.137 0.451 0.9988053 7776 tags=25%,

8907 75722 ^{1 1} list=35%, 1 signal=38%

Table 6R: GSEA for BAF180-null vs. BAF180-wildtype A704 cell lines on TNFA Founder gene sets

SCHOEN NFKB SIGNALING 33 0.4909 1.56925 0.01995 0.16486683 0.536 2415 tags=36%,

595 64 5654 list=ll%, signal=41%

AMIT SERUM RESPONSE 60 56 0.4424 1.57971 0.00460 0.17320979 0.494 2639 tags=27%,

MCFIOA 991 3 8295 list=12%,

signal=30%

MAHAJAN RESPONSE TO IL 72 0.4098 1.54688 0.01287 0.17401138 0.599 2709 tags=28%,

1A_UP 4586 62 5536 list=12%,

signal=32%

LINDSTEDT DENDRITIC CEL 58 0.4176 1.50155 0.01587 0.18031417 0.732 1356 tags=21%,

L MATURATION A 9278 03 3017 list=6%,

signal=22%

MEL 18 DN.V1 UP 135 0.3402 1.42544 0.01210 0.1834423 0.885 4045 tags=35%,

7582 44 6538 list=18%,

signal=42%

ALTEMEIER RESPONSE TO 107 0.3839 1.51305 0.00487 0.184769 0.706 2743 tags=25%,

LPS WITH MECHANICAL V 6505 21 8049 list=12%,

ENTILATION signal=29%

FERRARI RESPONSE TO FE 20 0.5746 1.61778 0.02637 0.1919037 0.406 1793 tags=30%,

NRETINIDE UP 8355 4 3627 list=8%,

signal=33%

BROWNE HCMV INFECTION 37 0.4298 1.43516 0.04749 0.19261208 0.874 2561 tags=19%,

2HR UP 0638 48 3402 list=12%,

signal=21%

BMI 1 DN MEL 18 DN.V 1_UP 139 0.3435 1.42645 0.00731 0.19325998 0.884 2687 tags=24%,

5652 67 7073 list=12%,

signal=28%

HINATA NFKB TARGETS K 85 0.3646 1.44192 0.02625 0.19637743 0.863 1759 tags=19%,

ERATINOCYTE UP 9486 37 8206 list=8%,

signal=20%

AMIT EGF RESPONSE 60 M 38 0.4603 1.51871 0.02643 0.1965755 0.69 1793 tags=21%,

CF10A 1594 85 1719 list=8%,

signal=23%

ZUCCHI METASTASIS DN 41 0.4910 1.64546 0.00223 0.19958329 0.33 1960 tags=20%,

1514 61 7137 list=9%,

signal=21%

GRAHAM CML QUIESCENT 21 0.4974 1.40506 0.09896 0.2010188 0.915 1356 tags=29%,

VS CML DIVIDING UP 8752 04 907 list=6%,

signal=30%

KRIEG HYPOXIA VIA KDM3 51 0.4483 1.58307 0.01716 0.2041567 0.488 2542 tags=27%,

A 681 49 7382 list=ll%, signal=31%

SEKI INFLAMMATORY RESP 73 0.3867 1.44430 0.02471 0.20557162 0.859 1356 tags=21%,

ONSE LPS UP 515 95 91 list=6%,

signal=22%

TIAN TNF SIGNALING VIA 28 0.4644 1.45253 0.05694 0.20914835 0.847 1356 tags=25%,

NFKB 4523 23 7608 list=6%,

signal=27%

MATTIOLI MGUS VS MULTI 16 0.5378 1.45509 0.06081 0.2228664 0.84 3175 tags=25%,

PLE MYELOMA 1915 39 081 list=14%,

signal=29%

DAZARD UV RESPONSE CL 19 0.4823 1.37395 0.12043 0.2244189 0.948 1356 tags=21%,

USTER G28 6227 26 0104 list=6%,

signal=22%

BURTON ADIPOGENESIS 1 33 0.4289 1.38067 0.06053 0.22504185 0.944 2940 tags=33%,

922 11 8117 list=13%,

signal=38%

UZONYI RESPONSE TO LEU 36 0.4034 1.34058 0.09071 0.2534478 0.974 1852 tags=17%,

KOTRIENE AND THROMBIN 385 54 274 list=8%,

signal=18%

MODULE 178 15 0.5160 1.33250 0.14516 0.25675952 0.981 2015 tags=33%,

862 39 129 list=9%,

signal=37%

HINATA NFKB TARGETS FI 80 0.3366 1.28551 0.09004 0.26094657 0.996 1759 tags=15%,

BROBLAST UP 1574 36 74 list=8%,

signal=16%

MCDOWELL ACUTE LUNG I 39 0.4124 1.34302 0.08163 0.26111743 0.974 1447 tags=18%,

NJURY UP 3193 83 265 list=7%,

signal=19%

BILD HRAS ONCOGENIC SI 240 0.2894 1.29007 0.03 0.2636602 0.994 2639 tags=17%,

GNATURE 971 07 list=12%,

signal=19% ALK_DN.V1_UP 113 0.3161 1.29322 0.03991 0.2682712 0.994 2807 tags=24%,

5773 28 1307 list=13%, signal=27%

KOBAYASHI EGFR SIGNALI 17 0.4668 1.27323 0.16339 0.27098557 0.996 4237 tags=35%,

NG 6HR DN 8193 17 87 list=19%, signal=44%

ZHOU INFLAMMATORY RES 442 0.2736 1.29449 0.01428 0.27551138 0.994 2640 tags=15%,

PONSE FIMA UP 1786 02 5714 list=12%, signal=17%

MODULE 362 19 0.4612 1.30058 0.13983 0.27594692 0.993 2015 tags=32%,

412 92 051 list=9%, signal=35%

WIEDERSCHAIN TARGETS 56 0.3694 1.31002 0.09032 0.28232476 0.99 2687 tags=23%,

OF BMI 1 AND PCGF2 282 4 2584 list=12%, signal=26%

GARGALOVIC RESPONSE T 33 0.4044 1.30250 0.11088 0.28378433 0.992 2730 tags=21%, o OXIDIZED PHOSPHOLIPID 8025 57 296 list=12%,

S BLACK UP signal=24%

RASHI NFKB 1 TAR GETS 18 0.4557 1.24937 0.19027 0.29257196 1356 tags=17%,

5842 24 483 ¹ list=6%, signal=18%

PLASARI TGFB1 TARGETS 1 188 0.2913 1.25345 0.05822 0.29428974 1 2815 tags=21%,

OHR UP 739 98 785 list=13%, signal=24%

P53 DN.V2 UP 117 0.2990 1.21592 0.11374 0.29627326 2730 tags=22%,

527 27 407 ¹ list=12%, signal=25%

BROCKE APOPTOSIS REVER 137 0.2968 1.24017 0.08395 0.29763865 1 2709 tags=19%,

SED BY IL6 2255 062 list=12%, signal=21%

BURTON ADIPOGENESIS PE 50 0.3495 1.21595 0.17050 0.30387002 1 2511 tags=22%,

AK AT 2HR 2435 03 691 list=ll%, signal=25%

SESTO RESPONSE TO UV C 20 0.4313 1.21871 0.21382 0.30712342 1 543 tags=15%,

3 2424 78 29 list=2%, signal=15%

SUZUKI RESPONSE TO TSA 19 0.4190 1.19375 0.22345 0.31074792 1 3262 tags=32%,

AND DECITABINE IA 3538 35 133 list=15%, signal=37%

THEILGAARD NEUTROPHIL 73 0.3150 1.19626 0.15311 0.31408814 1 1407 tags=ll%,

AT SKIN WOUND UP 8383 98 004 list=6%, signal=12%

DAZARD UV RESPONSE CL 29 0.3909 1.21938 0.19120 0.3144602 1 1856 tags=21%,

USTER G2 4698 48 88 list=8%, signal=23%

PHONG TNF RESPONSE NO 330 0.2594 1.19952 0.05851 0.3162518 1 3366 tags=21%,

T VIA P38 32 31 064 list=15%, signal=24%

HAHTOLA MYCOSIS FUNGO 58 0.3379 1.22101 0.16916 0.32014048 1 2059 tags=19%,

IDES CD4 UP 4093 18 488 list=9%, signal=21%

BMI1 DN.V1 UP 139 0.2809 1.18070 0.13711 0.32538497 1705 tags=17%,

1383 67 584 ¹ list=8%, signal=19%

ZWANG CLASS 3 TRANSIEN 206 0.2650 1.15335 0.13625 0.34272358 1 2516 tags=16%,

TLY INDUCED BY EGF 7318 95 866 list=ll%, signal=18%

GRAHAM CML QUIESCENT 50 0.3242 1.16590 0.21428 0.34277838 1 4211 tags=32%,

VS NORMAL DIVIDING UP 8753 62 572 list=19%, signal=39%

WANG TNF TARGETS 20 0.4111 1.15444 0.28051 0.34758896 1896 tags=20%,

3865 97 39 ¹ list=9%, signal=22%

GALINDO IMMUNE RESPON 79 0.3057 1.15475 0.20238 0.35437652 1 2059 tags=16%,

SE TO ENTEROTOXIN 5588 93 096 list=9%, signal=18%

ZHOU INFLAMMATORY RES 407 0.2400 1.13698 0.11653 0.35855886 1 2636 tags=15%,

PONSE LIVE UP 7683 1 116 list=12%, signal=17%

KIM WT l TARGETS UP 208 0.2550 1.13770 0.16707 0.36462373 2919 tags=17%,

9515 84 617 ¹ list=13%, signal=19% MODULE 516 16 0.4157 1.12728 0.31428 0.36917233 2015 tags=25%,

351 7 573 ¹ list=9%,

signal=27%

ZHOU INFLAMMATORY RES 342 0.2388 1.12140 0.13611 0.37299615 1 3023 tags=19%,

PONSE LPS UP 7469 55 111 list=14%,

signal=22%

AMIT EGF RESPONSE 40 HE 40 0.3365 1.10549 0.28854 0.3822015 1 1597 tags=15%,

LA 9357 82 626 list=7%,

signal=16%

BERENJENO TRANSFORMED 28 0.3586 1.10660 0.30997 0.38779154 1 871 tags=ll%,

BY RHOA FOREVER DN 3847 42 878 list=4%,

signal=ll%

ABE VEGFA TARGETS 30MI 24 0.3626 1.11012 0.30232 0.3884633 1 2516 tags=21%,

N 8952 85 558 list=ll%, signal=23%

WINZEN DEGRADED VIA K 97 0.2729 1.08726 0.28868 0.4096403 1 1356 tags=13%,

HSRP 0994 3 36 list=6%,

signal=14%

AMIT EGF RESPONSE 120 H 69 0.2867 1.06797 0.32558 0.41886824 1 1356 tags=12%,

ELA 648 13 14 list=6%,

signal=12%

RELA DN.V 1_UP 131 0.2548 1.06820 0.28078 0.42561394 2292 tags=13%,

3254 64 818 ¹ list=10%,

signal=14%

FOSTER TOLERANT MACRO 390 0.2248 1.06938 0.22762 0.43101433 1 3518 tags=17%,

PHAGE DN 4367 25 148 list=16%,

signal=20%

KIM WT1 TARGETS 12HR U 155 0.2512 1.06993 0.29128 0.43737483 1 1861 tags=14%,

P 566 76 44 list=8%,

signal=15%

DORN ADENOVIRUS INFEC 33 0.3322 1.05225 0.37938 0.4438031 1 2299 tags=15%,

TION 12HR DN 664 77 598 list=10%,

signal=17%

AMIT SERUM RESPONSE 40 30 0.3302 1.01804 0.42152 0.4886379 1 1356 tags=13%,

MCFIOA 3232 23 467 list=6%,

signal=14%

DIRMEIER LMP1 RESPONSE 62 0.2848 1.02832 0.38863 0.48976937 1 973 tags=10%,

EARLY 8106 98 635 list=4%,

signal=10%

TSAI RESPONSE TO IONIZI 142 0.2469 1.01853 0.40714 0.4951767 1 3391 tags=20%,

NG RADIATION 3018 73 285 list=15%,

signal=24%

SARTIPY BLUNTED BY INS 19 0.3673 1.02116 0.43064 0.49723047 1 5197 tags=53%,

ULIN RESISTANCE UP 268 6 183 list=23%,

signal=69%

AMIT DELAYED EARLY GE 18 0.3687 1.00762 0.43652 0.5051059 1 4689 tags=44%,

NES 8896 2 56 list=21%,

signal=56%

MODULE 444 17 0.3673 0.99056 0.47111 0.5356595 2015 tags=24%,

9218 363 112 ¹ list=9%,

signal=26%

KIM WT 1 TARGETS 8HR UP 160 0.2246 0.95971 0.54919 0.59616786 1977 tags=13%,

5761 507 91 ¹ list=9%,

signal=14%

OSWALD HEMATOPOIETIC 217 0.2076 0.91883 0.69873 0.6804433 1 2593 tags=12%,

STEM CELL IN COLLAGEN 2564 22 416 list=12%,

GEL UP signal=14%

ZHOU_TNF_SIGNALING_4HR 54 0.2586 0.90186 0.64755 0.7084501 2950 tags=13%,

7853 983 84 ¹ list=13%,

signal=15%

YAO TEMPORAL RESPONSE 26 0.2984 0.88950 0.62826 0.7239441 1 6098 tags=50%,

TO PROGESTERONE CLUST 7574 64 085 list=27%,

ER 5 signal=69%

CASORELLI ACUTE PROMY 160 0.2000 0.85706 0.82422 0.7685506 1 2726 tags=15%,

ELOCYTIC LEUKEMIA UP 1574 544 805 list=12%,

signal=17%

NEMETH INFLAMMATORY 83 0.2226 0.85983 0.76483 0.7736369 1 2701 tags=17%,

RESPONSE LPS UP 5787 57 52 list=12%,

signal=19%

DORN ADENOVIRUS INFEC 39 0.2445 0.80700 0.79223 0.8429193 1 2299 tags=13%,

TION 48HR DN 2195 11 746 list=10%,

signal=14% GESERICK TERT TARGETS 20 0.2554 0.73272 0.82993 0.9210089 1 2726 tags=15%, DN 6053 026 2 list=12%, signal=17%

ZHOU TNF SIGNALING 30M 52 0.2028 0.71236 0.94022 0.9262648 1 2321 tags=10%, IN 513 19 99 list=10%, signal=ll%

Table 6S: GSEA for BAF180-wildtype, BRGl-null vs. BAF180-wildtype, BRGl-wildtype

A704 cell lines on TNFA founder gene sets

ABE VEGFA TARGETS 16 0.24924 0.6367 0.92747 0.97078 1 886 tags=44%,

366 44 56 294 2 list=40%, signal=73%

RASHI RESPONSE TO IONIZING RADI 41 0.18592 0.5865 0.97368 0.97929 1 932 tags=46%, ATION 1 338 687 42 33 1 list=42%, signal=80%

Table 6T: GSEA for BAF180-null vs. BAF180-wildtype A704 cell lines on IL2 Founder gene sets

Table 6U: GSEA for BAF180-wildtype, BRGl-null vs. BAF180-wildtype, BRGl-wildtype

A704 cell lines on IL2 founder gene sets

Table 6V: Enriched GO terms for

KEGG CYTOKINE CYTOKINE RECEPTOR INTERACTION genes in BAF180- mutant GSEA enriched vs. BAF180-mutant GSEA depleted

Enriched genes = GSEA core enrichment (i.e. top ranked genes until running enrichment score hits peak)

Depleted genes = GSEA most negatively ranked genes (i.e. bottom ranked genes until running enrichment score hits trough)

system process : stmuus :

process (GO:0010557) eveopment :

Table 6W: GSEA results for gene sets enriched in pre-treatment patient tumors with truncating mutations in PBRM1

HALLMARK APOPTOSIS 160 0.26717153 1.5470705 0.004237 0.017733 0.09 9110 tags=34%,

288 185 list=25%, signal=45

%

HALLMARK IL2 STAT5 SIGNA 199 0.2597181 1.5294203 0.005555 0.018474 0.1 8984 tags=31%, LING 556 342 list=25%, signal=41

%

HALLMARK APICAL JUNCTIO 198 0.25475252 1.5265775 0 0.017752 0.10 7994 tags=28%, N 696 2 list=22%, signal=36

%

HALLMARK MYOGENESIS 200 0.25268936 1.5108361 0 0.018697 0.11 8694 tags=33%,

744 list=24%, signal=43

%

HALLMARK UV RESPONSE D 144 0.25802347 1.4569446 0 0.026672 0.16 8559 tags=35%, N 224 6 list=24%, signal=45

%

HALLMARK ESTROGEN RESP 200 0.23530972 1.4246706 0 0.032439 0.20 7210 tags=28%, ONSE EARLY 258 6 list=20%, signal=34

%

HALLMARK WNT BETA CATE 42 0.32717755 1.4034909 0.057324 0.038341 0.24 2943 tags=19%, NIN SIGNALING 84 142 9 list=8%, signal=21

%

HALLMARK HEDGEHOG SIGN 35 0.32948953 1.3634391 0.088235 0.048595 0.30 8388 tags=43%, ALING 3 615 6 list=23%, signal=56

%

HALLMARK ADIPOGENESIS 196 0.22999962 1.3584664 0.005376 0.048597 0.31 3003 tags=16%,

344 757 6 list=8%, signal=18

%

HALLMARK CHOLESTEROL H 73 0.26637354 1.3271515 0.068100 0.059607 0.38 1187 tags=49%, OMEOSTASIS 356 573 3 8 list=33%, signal=74

%

HALLMARK REACTIVE OXIGE 47 0.29103118 1.3039039 0.084302 0.069519 0.44 6860 tags=28%, N SPECIES PATHWAY 33 01 4 list=19%, signal=34

%

HALLMARK APICAL SURFACE 44 0.29906154 1.3030225 0.101190 0.067381 0.44 2182 tags=18%,

48 69 6 list=6%, signal=19

%

HALLMARK TGF BETA SIGNA 54 0.25017482 1.1652176 0.221875 0.195366 0.86 7986 tags=33%, LING 1 2 list=22%, signal=43

%

HALLMARK HEME METABOLI 196 0.19430974 1.1569836 0.129533 0.200190 0.87 7877 tags=27%, SM 68 72 6 list=22%, signal=34

%

HALLMARK PANCREAS BETA 38 0.26110435 1.1266103 0.263322 0.239128 0.92 2627 tags=16%, CELLS 9 23 6 list=7%, signal=17

%

HALLMARK FATTY ACID ME 158 0.18055953 1.0471649 0.348837 0.389121 0.98 8803 tags=35%, TABOLISM 2 1 8 list=25%, signal=46

%

HALLMARK COMPLEMENT 196 0.16831398 1.003772 0.461956 0.486448 0.99 8674 tags=28%,

53 38 5 list=24%, signal=37 %

HALLMARK UV RESPONSE UP 154 0.17562571 0.9861503 0.532019 0.517978 0.99 7933 tags=28%,

7 97 7 list=22%, signal=36

% HALLMARK KRAS SIGNALING 193 0.1584056 0.9371324 0.651428 0.639393 1 3260 tags=13%, DN 6 3 list=9%, signal=15

%

HALLMARK ANDROGEN RESP 99 0.1555059 0.8228370 0.875502 0.873910 1 8780 tags=27%, ONSE 5 6 list=24%, signal=36

%

Table 6X: GSEA results for gene sets enriched in pre-treatment patient tumors wildtype at

PBRM1

Incorporation by Reference

All publications, patents, and patent applications mentioned herein are hereby incorporated by reference in their entirety as if each individual publication, patent or patent application was specifically and individually indicated to be incorporated by reference. In case of conflict, the present application, including any definitions herein, will control.

Also incorporated by reference in their entirety are any polynucleotide and polypeptide sequences which reference an accession number correlating to an entry in a public database, such as those maintained by The Institute for Genomic Research (TIGR) on the world wide web and/or the National Center for Biotechnology Information (NCBI) on the world wide web.

Equivalents

Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the present invention described herein. Such equivalents are intended to be encompassed by the following claims.