SYSTEM OF PREDICTION OF RESPONSE TO CANCER THERAPY AND

METHODS OF USING THE SAME

RELATED APPLICATIONS

[0001] This application is a PCT Application claiming priority to United States

Provisional Application Serial No. 62/565,338, filed September 29, 2017, and United States Provisional Application Serial No. 62/719,906, August 20, 2018, both of which are incorporated by reference in their entireties.

FIELD OF INVENTION

[0002] The disclosure relates to methods and a system for predicting components of protein expression, or interrelated immune genes, the protein activity levels of such genes, which are used to establish a prognosis for a subject, predict the likelihood of a subject to respond to a therapy for treatment of a disease or disorder, and/or predict improved therapies for treatment of as disease or disorder.

BACKGROUND

[0003] Immune checkpoint blockade (ICB) therapy provides remarkable clinical gains, where melanoma is at the forefront of its success. However, only a subset of patients with advanced tumors currently benefit from these therapies, which at times incur considerable side-effects and costs.

SUMMARY OF EMBODIMENTS

[0004] Melanoma, even in its metastatic form, is one of a handful of cancers in which spontaneous regression has been frequently observed and has been tightly linked to immune response 7 ' 8. This led us to conjecture that the immune components governing spontaneous tumor regression may be a major determinant of immune responses to ICB. To this end, we focused on neuroblastoma (NB), where we could take advantage of an existing cohort of patients with transcriptomic and clinical outcome data. Interestingly, NB in children under 18 months of age manifests frequent spontaneous regression ⁹ that is mediated by cellular immunity, including tumor-infiltrating lymphocytes, tumor- targeted T-cells and anti-neural antibodies ¹⁰ . Moreover, NB is the first pediatric cancer with an FDA-approved immunotherapy (Dinutuximab), a monoclonal antibody targeting the disialoganglioside GD2 that is expressed in NB, melanoma, and other tumors 11 ' 12. We thus hypothesized that an immune-based predictor of NB spontaneous regression may effectively predict ICB response in melanoma.

[0005] In one aspect, the disclosure features a method of identifying antigen- specific immune activity in a subject or population of subjects, comprising (a) selecting, from the subject or the population diagnosed with cancer, at least a first pair of nucleic acid sequences comprising a first nucleic acid sequence and a second nucleic sequence, wherein the first nucleic acid sequence and the second nucleic acid sequence encode a immune checkpoint protein or variant thereof; (b) calculating the ratio of expression of the first nucleic acid sequence over the expression of the second nucleic acid sequence in a sample from the subject; (c) assigning a probability score to at least the first pair of nucleic acid sequences based upon the ratio of expression of at least the first pair of nucleic acid sequences; and (d) identifying antigen-specific immune activity of the subject or the population of subjects based upon the ratio of expression of the first and the second nucleic acid sequences. In one embodiment, the pair of nucleic acid sequences is chosen from two of the nucleic acid sequences that are at least 70% homologous to nucleic acid sequences encoding: VISTA, HVEM, PDL1, CD40, CD80, CD137L, BTLA, CD27, CD28, CD267, OX40L, CD86, CD200, CTLA4, CD200R, TTM3, and PD1. In one embodiment, the method further comprises the step of determining the average probability score over at least 5 pairs of nucleic acid sequences and wherein the step of predicting the subject or population's responsiveness to the therapy is based upon the average probability score assigned over the at least 5 pairs of nucleic acid sequences. In one embodiment, the method further comprises the step of determining the average probability score over at least 10 pairs of nucleic acid sequences and wherein the step of predicting the subject or population's responsiveness to the therapy is based upon the average probability score assigned over each of the at least 10 pairs of nucleic acid sequences. In one embodiment, the step of selecting at least one pair of nucleic acid sequences comprises selecting at least ten pairs of nucleic acid sequences, wherein the step of calculating the ratio of expression comprises calculating the ratio of mRNA or protein expression over each of the at least ten pairs of nucleic acid sequences, wherein the step of assigning a probability score comprises assigning a probability each of the a least ten pairs of nucleic acid sequences and subsequently calculating an average probability score, and wherein the step of identifying antigen- specific immune activity comprises comparing the average probability score to a threshold value of a probability score associated with a control subject, such that, if the average probability score of the subject is at or above the threshold value of a control subject, then the subject is characterized as having high antigen- specific immune activity; and, if the average probability score of the subject falls below the threshold value, the subject is characterized as not having a high antigen- specific immune activity. In one embodiment, the step of assigning a probability score further comprises (i) assigning a probability score of 1 if the ratio of expression of the at least one pair of nucleic acid sequences is at or above 1, (ii) assigning a probability score of 0 if the ratio of expression of the at least one pair of nucleic acid sequences is below 1; (iii) calculating an average probability score over each pair of nucleic acid sequences; and wherein the step of predicting antigen- specific immune activity further comprises comparing the average probability score to a threshold value, such that, if the average probability score is above the threshold value, then subject is characterized as having high antigen- specific immune activity; and, if the average probability score of the subject falls below the threshold value, the subject is characterized as not having a high antigen- specific immune activity. In one embodiment, the threshold value is from about 0.50 to about 0.95. In one embodiment, the threshold value is from about 0.69 to about 0.90. In one embodiment, the method further comprises a step of quantifying, or acquiring data comprising the quantity of, mRNA or protein expression of one or more nucleic acid sequences in a sample comprising one or a plurality of cancer cells. In one embodiment, the sample is a human tissue sample comprising a tissue from a brushing, biopsy, or surgical resection of a subject. In one embodiment, the sample comprises a cell that is freshly obtained, formalin fixed, alcohol-fixed and/or paraffin embedded. In one embodiment, the sample comprises one or more skin cells. In one embodiment, the sample comprises one or a plurality of cancer cells that are melanoma cells. In one embodiment, the step of identifying antigen- specific immune activity of the subject or the population of subjects based upon the ratio of expression of the first and the second nucleic acid sequences comprising comparing expression of at least the first pair of nucleic acid sequences in a subject or population of subjects diagnosed with cancer with expression of the first pair of nucleic acid sequences in a control subject or control population of subjects such that if the ratio of expression of the at least first pair of nucleic acid sequences in the subject is statistically higher than the ratio of expression of the first pair of nucleic acid sequences in the control subject, then the subject is characterized as having high level of antigen-specific immunity or an increased survival rate under immune checkpoint blockage therapy as compared to the control subject. In one embodiment, the step of comparing expression of at least the first pair of nucleic acid sequences in a subject or population of subjects with expression of the first pair of nucleic acid sequences in a control subject or control population of subjects comprises performing a Kaplan-Meier test using a two-sided log-rank test. In one embodiment, the step of calculating the ratio of expression of the first nucleic acid sequence over the expression of the second nucleic acid sequence in a sample from the subject comprises calculating the expression of each nucleic acid sequence with Formula:

1, expi(x) /exp _j (x); 0, otherwise,

wherein expi(x) and exp _j (x) denote the quantification of expression of genes i and j in the sample x; and wherein the expression of the first nucleic acid sequence i and the second nucleic acid sequence j are calculated by measuring RNA expression in the sample by immunohistochemistry, quantitative PCR, or fluorescence analysis.

[0006] In another aspect, the disclosure features a method of predicting responsiveness to immune checkpoint therapy of a subject or of a population of subjects diagnosed or suspected of having cancer comprising (a) selecting, from the subject or the population diagnosed with cancer, at least a first pair of nucleic acid sequences comprising a first nucleic acid sequence and a second nucleic sequence, wherein the first nucleic acid sequence and the second nucleic acid sequence encode a immune checkpoint protein or variant thereof and wherein the first or the second nucleic acid sequence encodes a protein or variant thereof associated with anti-immune checkpoint blockage therapy; (b) calculating the ratio of expression of the first nucleic acid sequence over the expression of the second nucleic acid sequence in a sample from the subject; (c) assigning a probability score to at least the first pair of nucleic acid sequences based upon the ratio of expression of at least the first pair of nucleic acid sequences; and (d) predicting responsiveness to the therapy in the subject or the population of subjects based upon the ratio of expression of the first and the second nucleic acid sequences. In one embodiment, the step of selecting at least one pair of nucleic acid sequences comprises selecting at least ten pairs of nucleic acid sequences, wherein the step of calculating the ratio of expression comprises calculating the ratio of mRNA or protein expression over each of the at least ten pairs of nucleic acid sequences, wherein the step of assigning a probability score comprises assigning a probability each of the a least ten pairs of nucleic acid sequences and subsequently calculating an average probability score, and wherein the step of predicting responsiveness to the therapy comprises comparing the average probability score to a threshold value of a probability score associated with a control subject, such that, if the average probability score of the subject is at or above the threshold value of a control subject, then the subject is characterized as being responsive to the therapy; and, if the average probability score of the subject falls below the threshold value, the subject is characterized as being non-responsive to the therapy. In one embodiment, the step of assigning a probability score further comprises (i) assigning a probability score of 1 if the ratio of expression of the at least one pair of nucleic acid sequences is at or above 1, (ii) assigning a probability score of 0 if the ratio of expression of the at least one pair of nucleic acid sequences is below 1; (iii) calculating an average probability score over each pair of nucleic acid sequences; and wherein the step of predicting the responsiveness to therapy further comprises comparing the average probability score to a threshold value, such that, if the average probability score is above the threshold value, then subject is characterized as being responsive to the therapy; and, of the average probability score falls below the threshold value, then the subject is characterized as being nonresponsive to the therapy. In one embodiment, the threshold value is from about 0.50 to about 0.95. In one embodiment, the threshold value is from about 0.69 to about 0.85. In one embodiment, the step of calculating the ratio of expression of the further comprises normalizing expression of the first pair of nucleic acid sequences by the ratio of expression of a pair of nucleic acid sequences unassociated with anti-CTLA-4 or anti-PDl blockade therapy used as a negative control. In one embodiment, the step of assigning a probability score comprises assigning a 1 to at least the first pair of nucleic acid sequences if the ratio of expression of the pair is 1 or greater and assigning a 0 to at least the first pair of nucleic acid sequences if the ratio of expression of the pair is less than 1. In one embodiment, the sample is a human tissue sample comprising a tissue from a brushing, biopsy, or surgical resection of a subject. In one embodiment, the sample comprises a cell that is freshly obtained, formalin fixed, alcohol-fixed and/or paraffin embedded. In one embodiment, the sample comprises one or more skin cells. In one embodiment, the sample comprises one or more cancer cells. In one embodiment, the cancer cell is a melanoma cell. In one embodiment, the subject is diagnosed or suspected of having melanoma. In one embodiment, the cancer is metastatic cancer derived from the skin. In one embodiment, the pair of nucleic acid sequences is chosen from two of the nucleic acid sequences that are at least 70% homologous to nucleic acid sequences encoding: VISTA, HVEM, PDL1, CD40, CD80, CD137L, BTLA, CD27, CD28, CD267, OX40L, CD86, CD200, CTLA4, CD200R, TTM3, and PD1. In one embodiment, the method further comprises the step of determining the average probability score over at least 5 pairs of nucleic acid sequences and wherein the step of predicting the subject or population's responsiveness to the therapy is based upon the average probability score assigned over the at least 5 pairs of nucleic acid sequences. In one embodiment, the method further comprises the step of determining the average probability score over at least 10 pairs of nucleic acid sequences and wherein the step of predicting the subject or population's responsiveness to the therapy is based upon the average probability score assigned over each of the at least 10 pairs of nucleic acid sequences. In one embodiment, the method further comprises the step of determining the average probability score over at least 15 pairs of nucleic acid sequences and wherein the step of predicting the subject or population's responsiveness to the therapy is based upon the average probability score assigned over each of the at least 15 pairs of nucleic acid sequences. In one embodiment, the method further comprises the step of acquiring mRNA or protein expression quantities from at least one sample of the subject. In one embodiment, the method further comprises quantifying the number of CD4+ cells and CD8+ cells in the sample. In one embodiment, the step of predicting responsiveness to the therapy comprises correlating expression of at least the first pair of nucleic acid sequences with antigen-specific immunity in the subject or the population of subjects comprises comparing expression of at least the first pair of nucleic acid sequences in a subject or population of subjects with expression of the first pair of nucleic acid sequences in a control subject or control population of subjects that have an low level of antigen-specific immunity or a decreased survival rate due to the cancer while taking the therapy. In one embodiment, the step of comparing expression of at least the first pair of nucleic acid sequences in a subject or population of subjects with expression of the first pair of nucleic acid sequences in a control subject or control population of subjects that have an low level of antigen- specific immunity or a decreased survival rate due to the cancer comprises performing a Kaplan-Meier test using a two-sided log-rank test. In one embodiment, the step of calculating the ratio of expression of the first nucleic acid sequence over the expression of the second nucleic acid sequence in a sample from the subject comprises calculating the expression of each nucleic acid sequence with Formula:

Fi _{, j} (x) = 1, expi(x) /exp _j (x); 0, otherwise,

wherein expi(x) and exp _j (x) denote the quantification of expression of genes i and j in the sample x. In one embodiment, the expression of genes i and j are calculated by measuring RNA expression in the sample by immunohistochemistry, quantitative RT-PCR, or microarray analysis. In one embodiment, the subject or population of subjects comprises data collected while the subject or population of subjects is exposed to therapy for no more than about 90 days. In one embodiment, the therapy is an antibody comprising a CDR sequence programmed death receptor- 1 (PD-1), an antibody comprising a CDR sequence that binds cytotoxic T lymphocyte-associated protein 4 (CTLA-4), or a combination thereof. In one embodiment, the method is a computer- implemented method, the method comprising: in a system configured to perform statistical analysis comprising at least one processor and a memory, performing statistical analysis or calculating a probability score of any of steps of the method. In one embodiment, the step of calculating the probability score or performing the statistical analysis, by the at least one processor, comprises: setting, by the at least one processor, a predetermined value, stored in the memory, that corresponds to a threshold value above which a nucleic acid sequence pair is correlated the immune activity or responsiveness of the subject to the therapy; calculating, by the at least one processor, the probability score, wherein calculating the probability score comprises receiving subject or population expression quantities of at least a first and second nucleic acid sequences in the at least one pair of nucleic acid sequences, normalizing the expression quantities against a control quantity, conducting one or a plurality of statistical tests from the expression quantities, and assigning a probability score based upon a comparison of an outcome of the statistical tests and the threshold value.

[0007] In another aspect, the disclosure features a method of predicting a prognosis and/or a clinical outcome of a subject or population of subjects suffering from cancer comprising: (a) selecting, from the subject or the population diagnosed with cancer, at least a first pair of nucleic acid sequences comprising a first nucleic acid sequence and a second nucleic sequence, wherein the first nucleic acid sequence and the second nucleic acid sequence encode a immune checkpoint protein or variant thereof; (b) calculating a ratio of expression of the first nucleic acid sequence over the expression of the second nucleic acid sequence in a sample from the subject; (c) assigning a probability score to at least the first pair of nucleic acid sequences based upon the ratio of expression of at least the first pair of nucleic acid sequences; and (d) predicting responsiveness to the therapy in the subject or the population of subjects based upon the ratio of expression of the first and the second nucleic acid sequences. In one embodiment, the step of selecting at least one pair of nucleic acid sequences comprises selecting at least ten pairs of nucleic acid sequences, wherein the step of calculating the ratio of expression comprises calculating the ratio of mRNA or protein expression over each of the at least ten pairs of nucleic acid sequences, wherein the step of assigning a probability score comprises assigning a probability each of the a least ten pairs of nucleic acid sequences and subsequently calculating an average probability score, and wherein the step of predicting responsiveness to the therapy comprises comparing the average probability score to a threshold value of a probability score associated with a control subject, such that, if the average probability score of the subject is at or above the threshold value of a control subject, then the subject is characterized as having a positive clinical outcome; and, if the average probability score of the subject falls below the threshold value, the subject is characterized as having a poor clinical outcome. In one embodiment, the step of assigning a probability score further comprises (i) assigning a probability score of 1 if the ratio of expression of the at least one pair of nucleic acid sequences is at or above 1, (ii) assigning a probability score of 0 if the ratio of expression of the at least one pair of nucleic acid sequences is below 1 ; (iii) calculating an average probability score over each pair of nucleic acid sequences; and wherein the step of predicting the responsiveness to therapy further comprises comparing the average probability score to a threshold value, such that, if the average probability score is above the threshold value, then subject is characterized as having a positive clinical outcome; and, of the average probability score falls below the threshold value, then the subject is characterized as having a poor clinical outcome. In one embodiment, the threshold value is from about 0.50 to about 0.95. In one embodiment, the threshold value is from about 0.69 to about 0.85. In one embodiment, the step of calculating the ratio of expression of the further comprises normalizing expression of the first pair of nucleic acid sequences by the ratio of expression of a pair of nucleic acid sequences unassociated with anti-CTLA-4 or anti-PDl blockade therapy used as a negative control. In one embodiment, the step of assigning a probability score comprises assigning a 1 to at least the first pair of nucleic acid sequences if the ratio of expression of the pair is 1 or greater and assigning a 0 to at least the first pair of nucleic acid sequences if the ratio of expression of the pair is less than 1. In one embodiment, the sample is a human tissue sample comprising a tissue from a brushing, biopsy, or surgical resection of a subject. In one embodiment, the sample comprises a cell that is freshly obtained, formalin fixed, alcohol-fixed and/or paraffin embedded. In one embodiment, the sample comprises one or more skin cells or cells derived form the skin. In one embodiment, the sample comprises one or more skin cancer cells. In one embodiment, the subject is diagnosed with melanoma. In one embodiment, the cancer is metastatic melanoma. In one embodiment, the pair of nucleic acid sequences is chosen from two of the nucleic acid sequences that are at least 70% homologous to nucleic acid sequences encoding: VISTA, HVEM, PDL1, CD40, CD80, CD137L, BTLA, CD27, CD28, CD267, OX40L, CD86, CD200, CTLA4, CD200R, TIM3, and PD1. In one embodiment, the method is a computer-implemented method, the method comprising: in a system configured to perform statistical analysis comprising at least one processor and a memory, performing statistical analysis or calculating a probability score of any of steps (a), (b), (c). In one embodiment, the step of calculating the probability score or performing the statistical analysis, by the at least one processor, comprises: setting, by the at least one processor, a threshold value, stored in the memory, that corresponds to a probability score above which the first pair of nucleic acid sequences is correlated to positive clinical outcome; calculating, by the at least one processor, the probability score, wherein calculating the probability score comprises analyzing a ratio of mRNA and/or protein expression associated with at least one pair of nucleic acid sequences associated with anti-immune checkpoint blockage therapy in a sample from a subject; conducting one or a plurality of statistical tests from the information associated with a disease or disorder; and assigning a probability score related to prognosis of the disease or disorder based upon a comparison of outcomes from the statistical tests and the threshold value.

[0008] In another aspect, the disclosure features a method of selecting or optimizing a therapy for treatment of a cancer responsive to immune checkpoint blockage therapy, the method comprising: (a) analyzing information from a subject or population of subjects associated with a disease or disorder comprising a step selecting at least a first pair of nucleic acids comprising a first and second nucleic acid; (i) wherein expression of the first nucleic acid sequence and second nucleic acid sequence contributes to antigen- specific immune response against one or more cancer cells; and/or (ii) wherein mRNA or protein expression of both nucleic acid sequences contributes at least partially to a clinical outcome more positive than control subject or control population; and (b) comparing expression of at least the first pair of nucleic acid sequences with a survival rate associated with a disease or disorder in a control population of subjects; and (c) assigning a probability score to the expression of at least the first pair of nucleic acid sequences based upon the survival rate of the subject or population of subjects diagnosed or suspected as having cancer; and (d) selecting a therapy useful for treatment of the disease or disorder based upon the expression of at least the first pair of nucleic acid sequences. In one embodiment, the step of comparing comprises determining the average probability score over at least 15 pairs of nucleic acid sequences and the step of predicting the subject or population's responsiveness to the therapy is based upon the average probability score assigned over each of the at least 15 pairs of nucleic acid sequences. In one embodiment, the method is a computer- implemented method, the method comprising: in a system configured to perform statistical analysis comprising at least one processor and a memory, performing statistical analysis or calculating a probability score of any of steps (a), (b), (c), and/or (d) by the processor. In one embodiment, the step of calculating the probability score or performing the statistical analysis, by the at least one processor, comprises: setting, by the at least one processor, a predetermined value, stored in the memory, that corresponds to a probability score above which the first pair of nucleic acid sequence is correlated to effectiveness of a therapy; calculating, by the at least one processor, the probability score, wherein calculating the probability score comprises analyzing information associated with a disease or disorder of the subject or the population of subjects; and conducting one or a plurality of statistical tests from the information associated with the cancer; and assigning a probability score related to effectiveness of a therapy based upon a comparison of outcomes from the statistical tests.

[0009] In another aspect, the disclosure features a computer program product encoded on a computer-readable storage medium comprising instructions for: (a) analyzing information from a subject or population of subjects associated with a disease or disorder comprising a step selecting at least a first pair of nucleic acids comprising a first and second nucleic acid; (i) wherein expression of the first nucleic acid sequence and second nucleic acid sequence contributes to antigen- specific immune response against one or more cancer cells; and/or (ii) wherein mRNA or protein expression of both nucleic acid sequences contributes at least partially to a clinical outcome more positive than control subject or control population; and (b) comparing a ratio of expression of at least the first pair of nucleic acid sequences with a survival rate associated with a disease or disorder in a control population of subjects; and (c) assigning a probability score to the expression of at least the first pair of nucleic acid sequences based upon the survival rate of the subject or population of subjects diagnosed with or suspected as having cancer; and (d) selecting a therapy useful for treatment of the disease or disorder based upon the expression of at least the first pair of nucleic acid sequences. In one embodiment, the computer program product further comprises instructions for: setting a threshold value that corresponds to a probability score above which the ratio of mRNA or protein expression of the first pair of nucleic acid sequence is correlated to effectiveness of treating a cancer; calculating the probability score, wherein calculating the probability score comprises analyzing information associated with a disease or disorder of the subject or the population of subjects; and conducting one or a plurality of statistical tests from the information associated with a disease or disorder; and assigning a probability score related to effectiveness or ineffectiveness of a therapy based upon a comparison of outcomes from the statistical tests. In one embodiment, the cancer is melanoma. In one embodiment, the therapy is immune checkpoint blockade therapy. In one embodiment, the computer program product further comprises the step of calculating the ratio of expression of the first nucleic acid sequence over the expression of the second nucleic acid sequence in a sample from the subject with Formula:

1, expi(x) /exp _j (x); 0, otherwise,

wherein expi(x) and exp _j (x) denote the quantification of expression of genes i and j in the sample x; and wherein the expression of the first nucleic acid sequence i and the second nucleic acid sequence j are calculated by measuring RNA expression in the sample by immunohistochemistry, quantitative PCR, or fluorescence analysis.

[0010] In another aspect, the disclosure features a system comprising a computer program product encoded on a computer-readable storage medium comprising instructions for: (a) analyzing information from a subject or population of subjects associated with a disease or disorder comprising a step selecting at least a first pair of nucleic acids comprising a first and second nucleic acid; (i) wherein expression of the first nucleic acid sequence and second nucleic acid sequence contributes to antigen- specific immune response against one or more cancer cells; and/or (ii) wherein mRNA or protein expression of both nucleic acid sequences contributes at least partially to a clinical outcome more positive than control subject or control population; and (b) comparing a ratio of expression of at least the first pair of nucleic acid sequences with a survival rate associated with a disease or disorder in a control population of subjects; and (c) assigning a probability score to the expression of at least the first pair of nucleic acid sequences based upon the survival rate of the subject or population of subjects diagnosed with or suspected as having cancer; and (d) selecting a therapy useful for treatment of the disease or disorder based upon the expression of at least the first pair of nucleic acid sequences. In one embodiment, the computer program product comprises web accessibility to a server in digital communication with a computer processor comprising mRNA or protein expression data from a cancer sample from a subject.

BRIEF DESCRIPTION OF DRAWINGS

[0011] FIG. 1A: Boxplots showing IMPRES of high vs low immune response in test and validation datasets of non-ICB treated melanoma patients ¹⁴ ; P-values are computed via a one-sided Rank-sum test. Boxplots center lines indicate medians, box edges represent the interquartile range, whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted individually using the '+' symbol. FIG. IB: Kaplan-Meier survival curves of patients with high versus low IMPRES (computed over the combined test and validation datasets ¹⁴ ). The median IMPRES is used to define the "Low IMPRES" and "High IMPRES" subgroups. The P-value is computed via a two-sided log-rank test. FIG. 1C: Upper Panel: Heatmaps showing the enrichment P-values for CDPs that are up (light gray) or down (dark gray) regulated in responders versus non-responders across the anti-PD-1 (encapsulated in the left rectangle) and the anti-CTLA-4 melanoma datasets ¹ ' ³ ' ⁴ ' ⁶ (right rectangle). The lower Panel displays the enrichment P-values for these CDPs in high immune response vs other subtypes in non-ICB treated melanoma, and in spontaneous regression vs non- spontaneous regression in the NB dataset. FIG. ID: Heatmaps showing the rank correlation p between expression levels of each CDP (vertical axis) and each of the IMPRES features ratios (horizontal axis), computed separately over the anti-PD-1 datasets, the anti-CTLA-4 datasets, the non-ICB treated melanoma datasets and the neuroblastoma dataset. White- colored entries denote non-statistically significant associations.

[0012] FIG. 2A: Receiver Operating Characteristic (ROC) curves quantifying

IMPRES prediction AUC across numerous publically available ICB response datasets ^{1 6} . FIG. 2B: ROC curves for the MGH dataset of ICB response (with 10 patients treated with anti-CTLA-4 and 31 patients treated with anti-PD-1) and for the aggregate datasets including all 297 samples, the 216 samples of patients treated with anti-PD-1 and 81 with anti-CTLA-4. FIG. 2C: Bar plots showing the prediction accuracy and error types for different IMPRES thresholds (where a positive label corresponds to a 'responder' prediction) on the aggregate compendium of 297 patients included in all 11 datasets studied. The dashed line represents the total number of responders. FIG. 2D: Precision/recall evaluation of IMPRES on the same aggregate compendium. The Y-axis displays the precision/recall as a function of the number of 'responder' predictions made (shown on the X-axis, obtained by decreasing the classification threshold, whose value is also displayed in italic font). Prediction performance in terms of specificity and sensitivity values is provided in Supp. Table 5. FIG. 2E-FIG. 2F: Kaplan Meier survival curves for the ICB treatment datasets ¹ ' ⁶ , with high vs. low IMPRES scores (using the median IMPRES as a threshold differentiating between the high and low groups). The P-values are computed via a two-sided log-rank test. FIG. 2G - FIG. 2H: Boxplots comparing progression free survival between low vs. high IMPRES in the ICB ¹ ' ⁵ datasets (using the median IMPRES as a differentiating threshold). P-values are computed via a one-sided Rank-sum test. Boxplots center lines indicate medians, box edges represent the interquartile range, whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted individually using the '+' symbol.

[0013] FIG. 3A: AUC of IMPRES and other published predictors across 9 publicly available ICB treatment datasets grouped by treatment type and stage (pre and on stands for before and during ICB treatment). The one-sided Rank-sum P-values comparing the performance of each predictor evaluated to that of IMPRES over all datasets are presented (P- value of 0.002 is achieved when IMPRES AUC is larger than that obtained by the other predictor for all 9 datasets, and 0.004 when it is larger for 8/9 datasets). Bar center is defined by the mean and error bars via SD. FIG. 3B: The empirical P-values comparing IMPRES performance to that of each of the other predictors in the three different ICB treatment classes and for the aggregate of all datasets (using n=1000 permutations, the value of '<le-3' denotes that IMPRES' prediction performance was superior to that of the predictor with which it was compared in all 1,000 repetitions). FIG. 3C: A network representation of the 15 pairwise features comprising IMPRES. Each node represents an immune checkpoint gene and each edge describes a pairwise relation (an IMPRES feature). The direction of edge A -> B denotes that the higher expression of A vs. that of B is associated with better patients' response. The color of the outline of each node denotes if it is inhibitory or activating and its fill color denotes whether it belongs to the PDl or CTLA-4 pathways. FIG. 3D: Clustogram (with average linkage function) of the individual predictive power of the 15 IMPRES features (based on their expression ratios) in each of the melanoma treatment datasets studied (the color scaling denotes the AUC obtained using each individual ratio as a response predictor, ranging from 0 to 1). FIG. 3E: Scatter plots showing the correlation between CIBERSORT- inferred CD8+ T cells abundance (X-axis) and the gene expression ratios of two IMPRES features that are significantly associated with it (Y-axis); CD40/PD1 (upper panel) and PD1/OX40L lower panel). The Spearman p and associated P-values are shown for each ICB response data ¹ ' ³ ' ⁴ ' ⁶ individually (on the right) and for all four datasets together (in the plot).

[0014] FIG. 4A: PCA analysis of the full transcriptomics of all NB samples (left panel: 176 that are clinically considered 'high risk' NB and 322 that are not; right panel: 181 that are clinically considered 'favorable disease course' NB (i.e. patients patient survived without chemotherapy for at least 1000 days) and 91 that are not (i.e., patient died despite intensive chemotherapy). FIG. 4B: Bar plot showing the AUCs for predicting spontaneous NB regression (Y-axis) resulting when selecting features for different score f binomial P- value thresholds (X-axis) when using the samples in the cluster defined by PC2+PC3>0 (dark gray) and when using all relevant samples (light gray). The number above each bar corresponds to the number of features selected with each P-value threshold. FIG. 4C: ROC curve depicting IMPRES predictive performance for predicting spontaneous regression on 108 NB samples including 92 spontaneously regressing patients and 16 progressing ones.

[0015] FIG. 5: Heatmap showing the fold change of CIBERSORT inferred immune cell abundances between NB patients with versus without spontaneous regression (row 1), ICB melanoma responders versus non-responders (row 2-5)^ and high immune response versus other subtypes in non-ICB treated melanoma patients (row 6) ⁵ . Entries bearing statistically significant differences include the corresponding P-values (using one-sided Rank-sum test). Four immune cell abundances are significantly up-regulated in regressing NB, from which 2 overlap with the 4 that are up-regulated in ICB responders in the Riaz et al. data (hyper-geometric P-value=0.01), 2 overlap with the 5 cell types that are up-regulated in ICB responders in the Van Allen et al. data (hyper-geometric P-value =0.023), and 1 overlaps with the single cell-type that is up-regulated in ICB responders in the Hugo et al. data (hyper- geometric P-value ~0). The overlap between the cell-types that are down-regulated in ICB responders/regressing NB is, however, not significant.

[0016] FIG. 6A and FIG. 6B are Precision/recall evaluation of IMPRES on the aggregate compendium of anti-PD-1 and anti-CTLA-4, respectively. The Y-axis displays the precision and recall of the response as a function of the number of 'responder' predictions made (shown on the X-axis, obtained by decreasing the classification threshold, whose value is also displayed in italic font). FIG. 6C: Survival prediction by IMPRES: Kaplan Meier survival curves for the Hugo et al6 data., using the median IMPRES score to define the Low and High groups, with the resulting one-sided log-rank P-value.

[0017] FIG. 7A: Scatter plots showing the correlation between CD8+, CD4+ T cells abundances (X-axis) inferred via CIBERSORT for different melanoma ICB response datasets (excluding nanostring datasets), and IMPRES scores (Y-axis). FIG. 7B: ROC curves denoting the accuracy of predicting ICB response from CIBERSORT inferred abundances of naive B cells, CD8+ and CD4+ T cells, for each of the RNA-seq melanoma ICB-treated datasets. FIG. 7C: Similar ICB response prediction ROC curves, but this time resulting from the most predictive CIBERSORT inferred abundances ratios.

[0018] FIG. 8A and FIG. 8D are Receiver Operating Characteristic (ROC) curves quantifying the prediction accuracy across numerous publicly available ICB response datasets 1 ·7· 2· 4· 8· 3, obtained via features (checkpoints binary ratios) selected analyzing the combined dataset from Hugo et al. and Van Allen et al. 1 ' 2 and the combined dataset from Riaz et al. and Hugo et al.2 ' 3 , respectively. FIG. 8B and FIG. 8E are boxplots showing the scores obtained via training on the combined data from Riaz et al. and Hugo et al. 2 ' 3 and on the combined dataset from Hugo et al and Van Allen et al. 1 ' 2 for high vs. low immune response metastatic melanoma (non ICB -treated) patients in published test and validation datasets ⁵ , compared via a one-sided rank-sum test. Boxplots centre lines indicate medians, box edges represent the interquartile range, whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted individually using the '+' symbol. FIG. 8C and FIG. 8F are Kaplan Meier survival curves for the combined datasets of non ICB -treated melanoma test and validation sets ⁵ , using the median of these inferred scores to define the Low and High groups, with the resulting log-rank P-values.

[0019] FIG. 9A: ROC curves quantifying the ICB response prediction AUCs obtained for numerous publicly available ICB response datasets 1 ·7· 2· 4· 8· 3, based on the reduced, 11 -features score (the features remaining after applying feature reduction to IMPRES features, listed in Supp. Table 9). FIG. 9B: ROC curves quantifying the ICB response prediction accuracy obtained for numerous publically available ICB response datasets 1—4 ' 7 ' 8 , based on features selected using the <18 months samples that are included in both PCA clusters (n=226 samples listed in Supp. Table 9).

[0020] FIG. 10A-FIG. IOC are boxplots comparing IMPRES scores of different melanoma subtypes (each subplot compares one subtype against all others, compared via two-sided rank-sum test) for the pre anti-PD-1, on anti-PD-1 and all samples from Riaz et al. . MU is mucosal melanoma; OC/OV is Ocular/Uveal melanoma; CU is Cutaneous melanoma and OT stands for Other subtypes'. Boxplots center lines indicate medians, box edges represent the interquartile range, whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted individually using the '+' symbol.

[0021] FIG. 11: Bar plot showing the number of randomly selected signatures (Y- axis) whose ICB response predictive power achieves AUCs higher than a given threshold (X- axis); the histogram presents the results for 1000 randomly generated predictors, each based on 15 immune gene binary relations that are randomly formed from the original space of 28 immune checkpoint genes (Supp. Table 1).

[0022] FIG. 12: Scatter plots visualizing the results of a PCA analysis using IMPRES ratios to describe each sample, for each ICB response data and for the integration of all data together. The variance explained by each PC is shown in the X, Y and Z-axes of each plot.

[0023] FIG. 13A and FIG. 13B are bar plots showing the mean and standard deviation of IMPRES scores (FIG. 13A) and mutation counts (FIG. 13B) for each cancer type, computed over the pertaining samples in the TCGA collection. FIG. 13C is a scatter plot showing the correlation between mean IMPRES scores (X-axis) and mean mutational counts (Y-axis) across the different cancer types (TCGA). The Spearman rank correlation between the two scores is almost 0.8. Circle size corresponds to sample size of each cancer type.

[0024] FIG. 14 is a flow chart showing the steps of the software method used to identify and process the nucleic acid sequence selected for assignment of a probability score correlated to the repsosiveness to the therapy by the subject.

DETAILED DESCRIPTION OF EMBODIMENTS

[0025] Various terms relating to the methods and other aspects of the present invention are used throughout the specification and claims. Such terms are to be given their ordinary meaning in the art unless otherwise indicated. Other specifically defined terms are to be construed in a manner consistent with the definition provided herein.

[0026] As used in this specification and the appended claims, the singular forms "a,"

"an," and "the" include plural referents unless the content clearly dictates otherwise.

[0027] The term "about" as used herein when referring to a measurable value such as an amount, a temporal duration, and the like, is meant to encompass variations of +20%, +10%, +5%, +1%, or +0.1% from the specified value, as such variations are appropriate to perform the disclosed methods.

[0028] The terms "amino acid" refer to a molecule containing both an amino group and a carboxyl group bound to a carbon which is designated the a-carbon. Suitable amino acids include, without limitation, both the D- and L-isomers of the naturally-occurring amino acids, as well as non-naturally occurring amino acids prepared by organic synthesis or other metabolic routes. In some embodiments, a single "amino acid" might have multiple sidechain moieties, as available per an extended aliphatic or aromatic backbone scaffold. Unless the context specifically indicates otherwise, the term amino acid, as used herein, is intended to include amino acid analogs including non-natural analogs.

[0029] As used herein, the terms "biopsy" means a cell sample, collection of cells, or bodily fluid removed from a subject or patient for analysis. In some embodiments, the biopsy is a bone marrow biopsy, punch biopsy, endoscopic biopsy, needle biopsy, shave biopsy, incisional biopsy, excisional biopsy, or surgical resection. [0030] As used herein, the terms "bodily fluid" means any fluid from isolated from a subject including, but not necessarily limited to, blood sample, serum sample, urine sample, mucus sample, saliva sample, and sweat sample. The sample may be obtained from a subject by any means such as intravenous puncture, biopsy, swab, capillary draw, lancet, needle aspiration, collection by simple capture of excreted fluid.

[0031] The terms "comprise(s)," "include(s)," "having," "has," "can," "contain(s)," and variants thereof, as used herein, are intended to be open-ended transitional phrases, terms, or words that do not preclude the possibility of additional acts or structures.

[0032] The terms "immune checkpoint blockade therapy" is a pharmaceutical composition comprising an antibody or antibody fragment that binds PD-1, CTLA-4 and/or a combination thereof.

[0033] The terms "immune checkpoint protein" is an amino acid sequence encoded by one or more nucleic acid sequences that activate the directly or indirectly effect the activation of the immune system within a subject. In some embodiments, the immune checkpoint protein comprises an amino acid sequence encoded by one of the following nucleic acid sequences or a nucleic acid variant that comprises at least about 70%, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to the following:

1561 gaagggtaca cagaaaaccc acagctcgaa gagtggtgac gtctggggtg gggaagaagg

1621 gtctggggg

PDL-1 1 gctttctatt caagtgcctt ctgtgtgtgc acatgtgtaa tacatatctg ggatcaaagc

(SEQ ID NO: 2) 61 tatctatata aagtccttga ttctgtgtgg gttcaaacac atttcaaagc ttcaggatcc

121 tgaaaggttt tgctctactt cctgaagacc tgaacaccgc tcccataaag ccatggcttg 181 ccttggattt cagcggcaca aggctcagct gaacctggct accaggacct ggccctgcac 241 tctcctgttt tttcttctct tcatccctgt cttctgcaaa gcaatgcacg tggcccagcc 301 tgctgtggta ctggccagca gccgaggcat cgccagcttt gtgtgtgagt atgcatctcc 361 aggcaaagcc actgaggtcc gggtgacagt gcttcggcag gctgacagcc aggtgactga 421 agtctgtgcg gcaacctaca tgatggggaa tgagttgacc ttcctagatg attccatctg 481 cacgggcacc tccagtggaa atcaagtgaa cctcactatc caaggactga gggccatgga 541 cacgggactc tacatctgca aggtggagct catgtaccca ccgccatact acctgggcat 601 aggcaacgga acccagattt atgtaattga tccagaaccg tgcccagatt ctgacttcct 661 cctctggatc cttgcagcag ttagttcggg gttgtttttt tatagctttc tcctcacagc 721 tgtttctttg agcaaaatgc taaagaaaag aagccctctt acaacagggg tctatgtgaa 781 aatgccccca acagagccag aatgtgaaaa gcaatttcag ccttatttta ttcccatcaa 841 ttgagaaacc attatgaaga agagagtcca tatttcaatt tccaagagct gaggcaattc 901 taactttttt gctatccagc tatttttatt tgtttgtgca tttgggggga attcatctct

961 ctttaatata aagttggatg cggaacccaa attacgtgta ctacaattta aagcaaagga 1021 gtagaaagac agagctggga tgtttctgtc acatcagctc cactttcagt gaaagcatca 1081 cttgggatta atatggggat gcagcattat gatgtgggtc aaggaattaa gttagggaat 1141 ggcacagccc aaagaaggaa aaggcaggga gcgagggaga agactatatt gtacacacct 1201 tatatttacg tatgagacgt ttatagccga aatgatcttt tcaagttaaa ttttatgcct 1261 tttatttctt aaacaaatgt atgattacat caaggcttca aaaatactca catggctatg 1321 ttttagccag tgatgctaaa ggttgtattg catatataca tatatatata tatatatata 1381 tatatatata tatatatata tatatatata tatattttaa tttgatagta ttgtgcatag

1441 agccacgtat gtttttgtgt atttgttaat ggtttgaata taaacactat atggcagtgt 1501 ctttccacct tgggtcccag ggaagttttg tggaggagct caggacacta atacaccagg 1561 tagaacacaa ggtcatttgc taactagctt ggaaactgga tgaggtcata gcagtgcttg 1621 attgcgtgga attgtgctga gttggtgttg acatgtgctt tggggctttt acaccagttc 1681 ctttcaatgg tttgcaagga agccacagct ggtggtatct gagttgactt gacagaacac 1741 tgtcttgaag acaatggctt actccaggag acccacaggt atgaccttct aggaagctcc 1801 agttcgatgg gcccaattct tacaaacatg tggttaatgc catggacaga agaaggcagc 1861 aggtggcaga atggggtgca tgaaggtttc tgaaaattaa cactgcttgt gtttttaact 1921 caatattttc catgaaaatg caacaacatg tataatattt ttaattaaat aaaaatctgt 1981 ggtggtcgtt ttccgga

CD28 1 taaagtcatc aaaacaacgt tatatcctgt gtgaaatgct gcagtcagga tgccttgtgg

(SEQ ID NO: 3) 61 tttgagtgcc ttgatcatgt gccctaaggg gatggtggcg gtggtggtgg ccgtggatga

121 cggagactct caggccttgg caggtgcgtc tttcagttcc cctcacactt cgggttcctc 181 ggggaggagg ggctggaacc ctagcccatc gtcaggacaa agatgctcag gctgctcttg 241 gctctcaact tattcccttc aattcaagta acaggaaaca agattttggt gaagcagtcg 301 cccatgcttg tagcgtacga caatgcggtc aaccttagct gcaagtattc ctacaatctc 361 ttctcaaggg agttccgggc atcccttcac aaaggactgg atagtgctgt ggaagtctgt 421 gttgtatatg ggaattactc ccagcagctt caggtttact caaaaacggg gttcaactgt 481 gatgggaaat tgggcaatga atcagtgaca ttctacctcc agaatttgta tgttaaccaa 541 acagatattt acttctgcaa aattgaagtt atgtatcctc ctccttacct agacaatgag 601 aagagcaatg gaaccattat ccatgtgaaa gggaaacacc tttgtccaag tcccctattt 661 cccggacctt ctaagccctt ttgggtgctg gtggtggttg gtggagtcct ggcttgctat 721 agcttgctag taacagtggc ctttattatt ttctgggtga ggagtaagag gagcaggctc 781 ctgcacagtg actacatgaa catgactccc cgccgccccg ggcccacccg caagcattac 841 cagccctatg ccccaccacg cgacttcgca gcctatcgct cctgacacgg acgcctatcc 901 agaagccagc cggctggcag cccccatctg ctcaatatca ctgctctgga taggaaatga 961 ccgccatctc cagccggcca cctcaggccc ctgttgggcc accaatgcca atttttctcg 1021 agtgactaga ccaaatatca agatcatttt gagactctga aatgaagtaa aagagatttc 1081 ctgtgacagg ccaagtctta cagtgccatg gcccacattc caacttacca tgtacttagt 1141 gacttgactg agaagttagg gtagaaaaca aaaagggagt ggattctggg agcctcttcc 1201 ctttctcact cacctgcaca tctcagtcaa gcaaagtgtg gtatccacag acattttagt 1261 tgcagaagaa aggctaggaa atcattcctt ttggttaaat gggtgtttaa tcttttggtt 1321 agtgggttaa acggggtaag ttagagtagg gggagggata ggaagacata tttaaaaacc 1381 attaaaacac tgtctcccac tcatgaaatg agccacgtag ttcctattta atgctgtttt 1441 cctttagttt agaaatacat agacattgtc ttttatgaat tctgatcata tttagtcatt 1501 ttgaccaaat gagggatttg gtcaaatgag ggattccctc aaagcaatat caggtaaacc 1561 aagttgcttt cctcactccc tgtcatgaga cttcagtgtt aatgttcaca atatactttc 1621 gaaagaataa aatagttctc ctacatgaag aaagaatatg tcaggaaata aggtcacttt 1681 atgtcaaaat tatttgagta ctatgggacc tggcgcagtg gctcatgctt gtaatcccag 1741 cactttggga ggccgaggtg ggcagatcac ttgagatcag gaccagcctg gtcaagatgg 1801 tgaaactccg tctgtactaa aaatacaaaa tttagcttgg cctggtggca ggcacctgta 1861 atcccagctg cccaagaggc tgaggcatga gaatcgcttg aacctggcag gcggaggttg 1921 cagtgagccg agatagtgcc acagctctcc agcctgggcg acagagtgag actccatctc 1981 aaacaacaac aacaacaaca acaacaacaa caaaccacaa aattatttga gtactgtgaa 2041 ggattatttg tctaacagtt cattccaatc agaccaggta ggagctttcc tgtttcatat 2101 gtttcagggt tgcacagttg gtctctttaa tgtcggtgtg gagatccaaa gtgggttgtg 2161 gaaagagcgt ccataggaga agtgagaata ctgtgaaaaa gggatgttag cattcattag 2221 agtatgagga tgagtcccaa gaaggttctt tggaaggagg acgaatagaa tggagtaatg 2281 aaattcttgc catgtgctga ggagatagcc agcattaggt gacaatcttc cagaagtggt 2341 caggcagaag gtgccctggt gagagctcct ttacagggac tttatgtggt ttagggctca 2401 gagctccaaa actctgggct cagctgctcc tgtaccttgg aggtccattc acatgggaaa 2461 gtattttgga atgtgtcttt tgaagagagc atcagagttc ttaagggact gggtaaggcc 2521 tgaccctgaa atgaccatgg atatttttct acctacagtt tgagtcaact agaatatgcc 2581 tggggacctt gaagaatggc ccttcagtgg ccctcaccat ttgttcatgc ttcagttaat 2641 tcaggtgttg aaggagctta ggttttagag gcacgtagac ttggttcaag tctcgttagt 2701 agttgaatag cctcaggcaa gtcactgccc acctaagatg atggttcttc aactataaaa 2761 tggagataat ggttacaaat gtctcttcct atagtataat ctccataagg gcatggccca 2821 agtctgtctt tgactctgcc tatccctgac atttagtagc atgcccgaca tacaatgtta 2881 gctattggta ttattgccat atagataaat tatgtataaa aattaaactg ggcaatagcc 2941 taagaagggg ggaatattgt aacacaaatt taaacccact acgcagggat gaggtgctat 3001 aatatgagga ccttttaact tccatcattt tcctgtttct tgaaatagtt tatcttgtaa 3061 tgaaatataa ggcacctccc acttttatgt atagaaagag gtcttttaat ttttttttaa 3121 tgtgagaagg aagggaggag taggaatctt gagattccag atcgaaaata ctgtactttg 3181 gttgattttt aagtgggctt ccattccatg gatttaatca gtcccaagaa gatcaaactc 3241 agcagtactt gggtgctgaa gaactgttgg atttaccctg gcacgtgtgc cacttgccag 3301 cttcttgggc acacagagtt cttcaatcca agttatcaga ttgtatttga aaatgacaga 3361 gctggagagt tttttgaaat ggcagtggca aataaataaa tacttttttt taaatggaaa 3421 gacttgatct atggtaataa atgattttgt tttctgactg gaaaaatagg cctactaaag 3481 atgaatcaca cttgagatgt ttcttactca ctctgcacag aaacaaagaa gaaatgttat 3541 acagggaagt ccgttttcac tattagtatg aaccaagaaa tggttcaaaa acagtggtag 3601 gagcaatgct ttcatagttt cagatatggt agttatgaag aaaacaatgt catttgctgc 3661 tattattgta agagtcttat aattaatggt actcctataa tttttgattg tgagctcacc 3721 tatttgggtt aagcatgcca atttaaagag accaagtgta tgtacattat gttctacata 3781 ttcagtgata aaattactaa actactatat gtctgcttta aatttgtact ttaatattgt 3841 cttttggtat taagaaagat atgctttcag aatagatatg cttcgctttg gcaaggaatt 3901 tggatagaac ttgctattta aaagaggtgt ggggtaaatc cttgtataaa tctccagttt 3961 agcctttttt gaaaaagcta gactttcaaa tactaatttc acttcaagca gggtacgttt 4021 ctggtttgtt tgcttgactt cagtcacaat ttcttatcag accaatggct gacctctttg 4081 agatgtcagg ctaggcttac ctatgtgttc tgtgtcatgt gaatgctgag aagtttgaca 4141 gagatccaac ttcagccttg accccatcag tccctcgggt taactaactg agccaccggt 4201 cctcatggct attttaatga gggtattgat ggttaaatgc atgtctgatc ccttatccca 4261 gccatttgca ctgccagctg ggaactatac cagacctgga tactgatccc aaagtgttaa 4321 attcaactac atgctggaga ttagagatgg tgccaataaa ggacccagaa ccaggatctt 4381 gattgctata gacttattaa taatccaggt caaagagagt gacacacact ctctcaagac 4441 ctggggtgag ggagtctgtg ttatctgcaa ggccatttga ggctcagaaa gtctctcttt 4501 cctatagata tatgcatact ttctgacata taggaatgta tcaggaatac tcaaccatca 4561 caggcatgtt cctacctcag ggcctttaca tgtcctgttt actctgtcta gaatgtcctt 4621 ctgtagatga cctggcttgc ctcgtcaccc ttcaggtcct tgctcaagtg tcatcttctc 4681 ccctagttaa actaccccac accctgtctg ctttccttgc ttatttttct ccatagcatt 4741 ttaccatctc ttacattaga catttttctt atttatttgt agtttataag cttcatgagg 4801 caagtaactt tgctttgttt cttgctgtat ctccagtgcc cagagcagtg cctggtatat 4861 aataaatatt tattgactga gtgaaaaaaa aaaaaaaaaa CD276 1 ccggcctcag ggacgcaccg gagccgcctt tccgggcctc aggcggattc tccggcgcgg

(SEQ ID NO: 4) 61 cccgccccgc ccctcggact ccccgggccg cccccggccc ccattcgggc cgggcctcgc

121 tgcggcggcg actgagccag gctgggccgc gtccctgagt cccagagtcg gcgcggcgcg 181 gcaggggcag ccttccacca cggggagccc agctgtcagc cgcctcacag gaagatgctg 241 cgtcggcggg gcagccctgg catgggtgtg catgtgggtg cagccctggg agcactgtgg 301 ttctgcctca caggagccct ggaggtccag gtccctgaag acccagtggt ggcactggtg 361 ggcaccgatg ccaccctgtg ctgctccttc tcccctgagc ctggcttcag cctggcacag 421 ctcaacctca tctggcagct gacagatacc aaacagctgg tgcacagctt tgctgagggc 481 caggaccagg gcagcgccta tgccaaccgc acggccctct tcccggacct gctggcacag 541 ggcaacgcat ccctgaggct gcagcgcgtg cgtgtggcgg acgagggcag cttcacctgc 601 ttcgtgagca tccgggattt cggcagcgct gccgtcagcc tgcaggtggc cgctccctac 661 tcgaagccca gcatgaccct ggagcccaac aaggacctgc ggccagggga cacggtgacc 721 atcacgtgct ccagctacca gggctaccct gaggctgagg tgttctggca ggatgggcag 781 ggtgtgcccc tgactggcaa cgtgaccacg tcgcagatgg ccaacgagca gggcttgttt 841 gatgtgcaca gcatcctgcg ggtggtgctg ggtgcaaatg gcacctacag ctgcctggtg 901 cgcaaccccg tgctgcagca ggatgcgcac agctctgtca ccatcacacc ccagagaagc 961 cccacaggag ccgtggaggt ccaggtccct gaggacccgg tggtggccct agtgggcacc 1021 gatgccaccc tgcgctgctc cttctccccc gagcctggct tcagcctggc acagctcaac 1081 ctcatctggc agctgacaga caccaaacag ctggtgcaca gtttcaccga aggccgggac 1141 cagggcagcg cctatgccaa ccgcacggcc ctcttcccgg acctgctggc acaaggcaat 1201 gcatccctga ggctgcagcg cgtgcgtgtg gcggacgagg gcagcttcac ctgcttcgtg 1261 agcatccggg atttcggcag cgctgccgtc agcctgcagg tggccgctcc ctactcgaag 1321 cccagcatga ccctggagcc caacaaggac ctgcggccag gggacacggt gaccatcacg 1381 tgctccagct accggggcta ccctgaggct gaggtgttct ggcaggatgg gcagggtgtg 1441 cccctgactg gcaacgtgac cacgtcgcag atggccaacg agcagggctt gtttgatgtg 1501 cacagcgtcc tgcgggtggt gctgggtgcg aatggcacct acagctgcct ggtgcgcaac 1561 cccgtgctgc agcaggatgc gcacggctct gtcaccatca cagggcagcc tatgacattc 1621 cccccagagg ccctgtgggt gaccgtgggg ctgtctgtct gtctcattgc actgctggtg 1681 gccctggctt tcgtgtgctg gagaaagatc aaacagagct gtgaggagga gaatgcagga 1741 gctgaggacc aggatgggga gggagaaggc tccaagacag ccctgcagcc tctgaaacac 1801 tctgacagca aagaagatga tggacaagaa atagcctgac catgaggacc agggagctgc 1861 tacccctccc tacagctcct accctctggc tgcaatgggg ctgcactgtg agccctgccc 1921 ccaacagatg catcctgctc tgacaggtgg gctccttctc caaaggatgc gatacacaga 1981 ccactgtgca gccttatttc tccaatggac atgattccca agtcatcctg ctgccttttt 2041 tcttatagac acaatgaaca gaccacccac aaccttagtt ctctaagtca tcctgcctgc 2101 tgccttattt cacagtacat acatttctta gggacacagt acactgacca catcaccacc 2161 ctcttcttcc agtgctgcgt ggaccatctg gctgcctttt ttctccaaaa gatgcaatat 2221 tcagactgac tgaccccctg ccttatttca ccaaagacac gatgcatagt caccccggcc 2281 ttgtttctcc aatggccgtg atacactagt gatcatgttc agccctgctt ccacctgcat 2341 agaatctttt cttctcagac agggacagtg cggcctcaac atctcctgga gtctagaagc 2401 tgtttccttt cccctccttc ctcctcttgc tctagcctta atactggcct tttccctccc 2461 tgccccaagt gaagacaggg cactctgcgc ccaccacatg cacagctgtg catggagacc 2521 tgcaggtgca cgtgctggaa cacgtgtggt tcccccctgg cccagcctcc tctgcagtgc 2581 ccctctcccc tgcccatcct ccccacggaa gcatgtgctg gtcacactgg ttctccaggg 2641 gtctgtgatg gggcccctgg gggtcagctt ctgtccctct gccttctcac ctctttgttc 2701 ctttcttttc atgtatccat tcagttgatg tttattgagc aactacagat gtcagcactg 2761 tgttaggtgc tgggggccct gcgtgggaag ataaagttcc tccctcaagg actccccatc 2821 cagctgggag acagacaact aactacactg caccctgcgg tttgcagggg gctcctgcct 2881 ggctccctgc tccacacctc ctctgtggct caaggcttcc tggatacctc acccccatcc 2941 cacccataat tcttacccag agcatggggt tggggcggaa acctggagag agggacatag 3001 cccctcgcca cggctagaga atctggtggt gtccaaaatg tctgtccagg tgtgggcagg 3061 tgggcaggca ccaaggccct ctggaccttt catagcagca gaaaaggcag agcctggggc 3121 agggcagggc caggaatgct ttggggacac cgaggggact gccccccacc cccaccatgg 3181 tgctattctg gggctggggc agtcttttcc tggcttgcct ctggccagct cctggcctct 3241 ggtagagtga gacttcagac gttctgatgc cttccggatg tcatctctcc ctgccccagg 3301 aatggaagat gtgaggactt ctaatttaaa tgtgggactc ggagggattt tgtaaactgg 3361 gggtatattt tggggaaaat aaatgtcttt gtaaaaagct taaaaaaaaa aaaaaaaaa

CD80 1 gacaagtact gagtgaactc aaaccctctg taaagtaaca gaagttagaa ggggaaatgt (SEQ ID NO: 5) 61 cgcctctctg aagattaccc aaagaaaaag tgatttgtca ttgctttata gactgtaaga

121 agagaacatc tcagaagtgg agtcttaccc tgaaatcaaa ggatttaaag aaaaagtgga 181 atttttcttc agcaagctgt gaaactaaat ccacaacctt tggagaccca ggaacaccct

241 ccaatctctg tgtgttttgt aaacatcact ggagggtctt ctacgtgagc aattggattg

301 tcatcagccc tgcctgtttt gcacctggga agtgccctgg tcttacttgg gtccaaattg

361 ttggctttca cttttgaccc taagcatctg aagccatggg ccacacacgg aggcagggaa

421 catcaccatc caagtgtcca tacctcaatt tctttcagct cttggtgctg gctggtcttt

481 ctcacttctg ttcaggtgtt atccacgtga ccaaggaagt gaaagaagtg gcaacgctgt

541 cctgtggtca caatgtttct gttgaagagc tggcacaaac tcgcatctac tggcaaaagg

601 agaagaaaat ggtgctgact atgatgtctg gggacatgaa tatatggccc gagtacaaga

661 accggaccat ctttgatatc actaataacc tctccattgt gatcctggct ctgcgcccat

721 ctgacgaggg cacatacgag tgtgttgttc tgaagtatga aaaagacgct ttcaagcggg

781 aacacctggc tgaagtgacg ttatcagtca aagctgactt ccctacacct agtatatctg

841 actttgaaat tccaacttct aatattagaa ggataatttg ctcaacctct ggaggttttc

901 cagagcctca cctctcctgg ttggaaaatg gagaagaatt aaatgccatc aacacaacag

961 tttcccaaga tcctgaaact gagctctatg ctgttagcag caaactggat ttcaatatga

1021 caaccaacca cagcttcatg tgtctcatca agtatggaca tttaagagtg aatcagacct

1081 tcaactggaa tacaaccaag caagagcatt ttcctgataa cctgctccca tcctgggcca

1141 ttaccttaat ctcagtaaat ggaatttttg tgatatgctg cctgacctac tgctttgccc

1201 caagatgcag agagagaagg aggaatgaga gattgagaag ggaaagtgta cgccctgtat

1261 aacagtgtcc gcagaagcaa ggggctgaaa agatctgaag gtcccacctc catttgcaat

1321 tgacctcttc tgggaacttc ctcagatgga caagattacc ccaccttgcc ctttacgtat

1381 ctgctcttag gtgcttcttc acttcagttg ctttgcagga agtgtctaga ggaatatggt

1441 gggcacagaa gtagctctgg tgaccttgat caaggtgttt tgaaatgcag aattcttgag

1501 ttctggaagg gactttagag aataccagtg ttattaatga caaaggcact gaggcccagg

1561 gaggtgaccc gaattataaa ggccagcgcc agaacccaga tttcctaact ctggtgctct

1621 ttccctttat cagtttgact gtggcctgtt aactggtata tacatatata tgtcaggcaa

1681 agtgctgctg gaagtagaat ttgtccaata acaggtcaac ttcagagact atctgatttc

1741 ctaatgtcag agtagaagat tttatgctgc tgtttacaaa agcccaatgt aatgcatagg

1801 aagtatggca tgaacatctt taggagacta atggaaatat tattggtgtt tacccagtat

1861 tccatttttt tcattgtgtt ctctattgct gctctctcac tcccccatga ggtacagcag

1921 aaaggagaac tatccaaaac taatttcctc tgacatgtaa gacgaatgat ttaggtacgt

1981 caaagcagta gtcaaggagg aaagggatag tccaaagact taactggttc atattggact

2041 gataatctct ttaaatggct ttatgctagt ttgacctcat ttgtaaaata tttatgagaa

2101 agttctcatt taaaatgaga tcgttgttta cagtgtatgt actaagcagt aagctatctt

2161 caaatgtcta aggtagtaac tttccatagg gcctccttag atccctaaga tggctttttc

2221 tccttggtat ttctgggtct ttctgacatc agcagagaac tggaaagaca tagccaactg

2281 ctgttcatgt tactcatgac tcctttctct aaaactgcct tccacaattc actagaccag

2341 aagtggacgc aacttaagct gggataatca cattatcatc tgaaaatctg gagttgaaca

2401 gcaaaagaag acaacatttc tcaaatgcac atctcatggc agctaagcca catggctggg

2461 atttaaagcc tttagagcca gcccatggct ttagctacct cactatgctg cttcacaaac

2521 cttgctcctg tgtaaaacta tattctcagt gtagggcaga gaggtctaac accaacataa

2581 ggtactagca gtgtttcccg tattgacagg aatacttaac tcaataattc ttttcttttc

2641 catttagtaa cagttgtgat gactatgttt ctattctaag taattcctgt attctacagc

2701 agatactttg tcagcaatac taagggaaga aacaaagttg aaccgtttct ttaataa

HVEM 1 tccttcatac cggcccttcc cctcggcttt gcctggacag ctcctgcctc ccgcagggcc

(SEQ ID NO: 6) 61 cacctgtgtc ccccagcgcc gctccaccca gcaggcctga gcccctctct gctgccagac

121 accccctgct gcccactctc ctgctgctcg ggttctgagg cacagcttgt cacaccgagg

181 cggattctct ttctctttct ctttctcttc tggcccacag ccgcagcaat ggcgctgagt

241 tcctctgctg gagttcatcc tgctagctgg gttcccgagc tgccggtctg agcctgaggc

301 atggagcctc ctggagactg ggggcctcct ccctggagat ccacccccaa aaccgacgtc

361 ttgaggctgg tgctgtatct caccttcctg ggagccccct gctacgcccc agctctgccg

421 tcctgcaagg aggacgagta cccagtgggc tccgagtgct gccccaagtg cagtccaggt

481 tatcgtgtga aggaggcctg cggggagctg acgggcacag tgtgtgaacc ctgccctcca

541 ggcacctaca ttgcccacct caatggccta agcaagtgtc tgcagtgcca aatgtgtgac

601 ccagccatgg gcctgcgcgc gagccggaac tgctccagga cagagaacgc cgtgtgtggc

661 tgcagcccag gccacttctg catcgtccag gacggggacc actgcgccgc gtgccgcgct

721 tacgccacct ccagcccggg ccagagggtg cagaagggag gcaccgagag tcaggacacc

781 ctgtgtcaga actgcccccc ggggaccttc tctcccaatg ggaccctgga ggaatgtcag

841 caccagacca agtgcagctg gctggtgacg aaggccggag ctgggaccag cagctcccac

901 tgggtatggt ggtttctctc agggagcctc gtcatcgtca ttgtttgctc cacagttggc

961 ctaatcatat gtgtgaaaag aagaaagcca aggggtgatg tagtcaaggt gatcgtctcc 1021 gtccagcgga aaagacagga ggcagaaggt gaggccacag tcattgaggc cctgcaggcc

1081 cctccggacg tcaccacggt ggccgtggag gagacaatac cctcattcac ggggaggagc

1141 ccaaaccact gacccacaga ctctgcaccc cgacgccaga gatacctgga gcgacggctg

1201 ctgaaagagg ctgtccacct ggcggaacca ccggagcccg gaggcttggg ggctccgccc

1261 tgggctggct tccgtctcct ccagtggagg gagaggtggg gcccctgctg gggtagagct

1321 ggggacgcca cgtgccattc ccatgggcca gtgagggcct ggggcctctg ttctgctgtg

1381 gcctgagctc cccagagtcc tgaggaggag cgccagttgc ccctcgctca cagaccacac

1441 acccagccct cctgggccag cccagagggc ccttcagacc ccagctgtct gcgcgtctga

1501 ctcttgtggc ctcagcagga caggccccgg gcactgcctc acagccaagg ctggactggg

1561 ttggctgcag tgtggtgttt agtggatacc acatcggaag tgattttcta aattggattt

1621 gaattcggct cctgttttct atttgtcatg aaacagtgta tttggggaga tgctgtggga

1681 ggatgtaaat atcttgtttc tcctcaaact gtcacctccc ggtgtttctt gctgaacaag

1741 gagttccagg atggctgctg ggctgttcgg gggacccctg ccctcctccc gtcatgcctg

1801 ggggttcact ccacccagag aggagccctg gccgcccctt catatcccaa cagctgagct

1861 ctcagtgggc tcttctgacc tctgtggctc cgtccgaggc tattgctgtg gattctgatg

1921 ctcaaatggt gtcagatttg cccagtaaaa accccagatc tacatctgac ctacacttcc

1981 cagctgtgtc caccgagaaa ccccagtatc agtgacgcct gctgtgccca gccctctcca

2041 cctgctccgg gaacccgcca ggcccaggtc ccgctggcag gggcttcacc aggcctctga

2101 gccacacatt catttaatgg tcgggatgag gcccctttcc ccacatctga agttagaagc

2161 ggtgagggga atgaccctgc agccatgcca tgaggatgga ggccacatag cccctccgag

2221 catgcccgct ccaccccgcc ctaccccctc tcctttcctt gtcacctgcc tccagcagag

2281 cccccaggct gagccaccca ccccaactcc tctcctgcca ccccttgtcc tgtggaagct

2341 ttggcttagc gtcctggggt gtggagaggc ccatgcaggc caggtggagc cctgggcccc

2401 tagaaagcag cacttctggc tgccccaccc cgtgtcaccc tctccccaac tggaggcgtg

2461 gtctccaggg accacgggcc tccctgtgca tggaccggct cctgaccacc gtccagggtc

2521 attgccaggg taccttttca gaggctgacc ccatagacct ggctgccccc cagtgctaga

2581 tgggagccaa gcacagcctg cccttctgcc cacagtcccg ggggcaggtg ggagcatggg

2641 gccatggagt gagcgggcag gggtggcaga gggctccctg gtcaggggcc ccaacttccc

2701 ttcccccagg gaggccacct gacatctggg ctccaggcac agcaggaagc ccacctgccc

2761 caacctgtag ctcctcctcc tgggaggagc catggatcct ggaaaagctc tggggccacc

2821 tcccaggttt ggggggacag agctccaaga gacgacggct ggggacacga gccctcatgg

2881 ggccgctgtg tgctcacccc ttgattttct tcttttcatg catgagatta ggccaagtgt

2941 ggagaaatca atgatgttga cgatgaggct ccctgagaga aatcacaccc agcgggagct

3001 gctgctccca ggtctggcct cggtcaccag ccacctgctg catccgcggg agtggggccg

3061 aggacatggg agtggcaggt gcagcccccg gtactcactc agccccaggg agtgtccctg

3121 gctcccaggg ctctgggagg tgagggcagg tcccggggga ggctgggtta gtggcagctc

3181 cgggatgaga cctcagaggt ctgtctgact tgtccaagcc cggctatggg gaggtggggg

3241 gaaggaagga agaggagaga aataaggaga ggctgggcaa agaagacagg acggcagagg

3301 gagaggggag agaagtggga ggcagccagc agcgcagggc cctgagagta tttcagcggc

3361 accgctgtcc tgggccgccc ggtgccacat ctttgaaaac agttgtttaa tttaagcttg

3421 tccactcagt agctgttgaa tgtgggaggt tatcttgttc tattcaagtt gctataaaaa

3481 taaaaactac catagactgg gaaaaaaaaa aaaaaaaaa

CD86 1 ¾ 'tcattgcc gaggaaggct tgcacagggt gaaagctttg cttctctgct gctgtaacag

(SEQ ID NO: 7) 61 ggactagcac agacacacgg atgagtgggg tcatttccag atattaggtc acagcagaag

121 cagccaaaat ggatccccag tgcactatgg gactgagtaa cattctcttt gtgatggcct

181 tcctgctctc tggtgctgct cctctgaaga ttcaagctta tttcaatgag actgcagacc

241 tgccatgcca atttgcaaac tctcaaaacc aaagcctgag tgagctagta gtattttggc

301 aggaccagga aaacttggtt ctgaatgagg tatacttagg caaagagaaa tttgacagtg

361 ttcattccaa gtatatgggc cgcacaagtt ttgattcgga cagttggacc ctgagacttc

421 acaatcttca gatcaaggac aagggcttgt atcaatgtat catccatcac aaaaagccca

481 caggaatgat tcgcatccac cagatgaatt ctgaactgtc agtgcttgct aacttcagtc

541 aacctgaaat agtaccaatt tctaatataa cagaaaatgt gtacataaat ttgacctgct

601 catctataca cggttaccca gaacctaaga agatgagtgt tttgctaaga accaagaatt

661 caactatcga gtatgatggt attatgcaga aatctcaaga taatgtcaca gaactgtacg

721 acgtttccat cagcttgtct gtttcattcc ctgatgttac gagcaatatg accatcttct

781 gtattctgga aactgacaag acgcggcttt tatcttcacc tttctctata gagcttgagg

841 accctcagcc tcccccagac cacattcctt ggattacagc tgtacttcca acagttatta

901 tatgtgtgat ggttttctgt ctaattctat ggaaatggaa gaagaagaag cggcctcgca

961 actcttataa atgtggaacc aacacaatgg agagggaaga gagtgaacag accaagaaaa

1021 gagaaaaaat ccatatacct gaaagatctg atgaagccca gcgtgttttt aaaagttcga 1081 agacatcttc atgcgacaaa agtgatacat gtttttaatt aaagagtaaa gcccatacaa

1141 gtattcattt tttctaccct ttcctttgta agttcctggg caaccttttt gatttcttcc

1201 agaaggcaaa aagacattac catgagtaat aagggggctc caggactccc tctaagtgga 1261 atagcctccc tgtaactcca gctctgctcc gtatgccaag aggagacttt aattctctta 1321 ctgcttcttt tcacttcaga gcacacttat gggccaagcc cagcttaatg gctcatgacc 1381 tggaaataaa atttaggacc aatacctcct ccagatcaga ttcttctctt aatttcatag 1441 attgtgtttt ttttttaaat agacctctca atttctggaa aactgccttt tatctgccca

1501 gaattctaag ctggtgcccc actgaatttt gtgtacctgt gactaaacaa ctacctcctc 1561 agtctgggtg ggacttatgt atttatgacc ttatagtgtt aatatcttga aacatagaga 1621 tctatgtact gtaatagtgt gattactatg ctctagagaa aagtctaccc ctgctaagga 1681 gttctcatcc ctctgtcagg gtcagtaagg aaaacggtgg cctagggtac aggcaacaat 1741 gagcagacca acctaaattt ggggaaatta ggagaggcag agatagaacc tggagccact 1801 tctatctggg ctgttgctaa tattgaggag gcttgcccca cccaacaagc catagtggag 1861 agaactgaat aaacaggaaa atgccagagc ttgtgaaccc tgtttctctt gaagaactga 1921 ctagtgagat ggcctgggga agctgtgaaa gaaccaaaag agatcacaat actcaaaaga 1981 gagagagaga gaaaaaagag agatcttgat ccacagaaat acatgaaatg tctggtctgt 2041 ccaccccatc aacaagtctt gaaacaagca acagatggat agtctgtcca aatggacata 2101 agacagacag cagtttccct ggtggtcagg gaggggtttt ggtgataccc aagttattgg 2161 gatgtcatct tcctggaagc agagctgggg agggagagcc atcaccttga taatgggatg 2221 aatggaagga ggcttaggac tttccactcc tggctgagag aggaagagct gcaacggaat 2281 taggaagacc aagacacaga tcacccgggg cttacttagc ctacagatgt cctacgggaa 2341 cgtgggctgg cccagcatag ggctagcaaa tttgagttgg atgattgttt ttgctcaagg 2401 caaccagagg aaacttgcat acagagacag atatactggg agaaatgact ttgaaaacct 2461 ggctctaagg tgggatcact aagggatggg gcagtctctg cccaaacata aagagaactc 2521 tggggagcct gagccacaaa aatgttcctt tattttatgt aaaccctcaa gggttataga 2581 ctgccatgct agacaagctt gtccatgtaa tattcccatg tttttaccct gcccctgcct 2641 tgattagact cctagcacct ggctagtttc taacatgttt tgtgcagcac agtttttaat 2701 aaatgcttgt tacattcatt taaaaaaaaa aaaaa

CD27 1 cggaagggga agggggtgga ggttgctgct atgagagaga aaaaaaaaac agccacaata

(SEQ ID NO: 8) 61 gagattctgc cttcaaaggt tggcttgcca cctgaagcag ccactgccca gggggtgcaa

121 agaagagaca gcagcgccca gcttggaggt gctaactcca gaggccagca tcagcaactg 181 ggcacagaaa ggagccgcct gggcagggac catggcacgg ccacatccct ggtggctgtg 241 cgttctgggg accctggtgg ggctctcagc tactccagcc cccaagagct gcccagagag 301 gcactactgg gctcagggaa agctgtgctg ccagatgtgt gagccaggaa cattcctcgt 361 gaaggactgt gaccagcata gaaaggctgc tcagtgtgat ccttgcatac cgggggtctc 421 cttctctcct gaccaccaca cccggcccca ctgtgagagc tgtcggcact gtaactctgg 481 tcttctcgtt cgcaactgca ccatcactgc caatgctgag tgtgcctgtc gcaatggctg 541 gcagtgcagg gacaaggagt gcaccgagtg tgatcctctt ccaaaccctt cgctgaccgc 601 tcggtcgtct caggccctga gcccacaccc tcagcccacc cacttacctt atgtcagtga 661 gatgctggag gccaggacag ctgggcacat gcagactctg gctgacttca ggcagctgcc 721 tgcccggact ctctctaccc actggccacc ccaaagatcc ctgtgcagct ccgattttat 781 tcgcatcctt gtgatcttct ctggaatgtt ccttgttttc accctggccg gggccctgtt 841 cctccatcaa cgaaggaaat atagatcaaa caaaggagaa agtcctgtgg agcctgcaga 901 gccttgtcgt tacagctgcc ccagggagga ggagggcagc accatcccca tccaggagga 961 ttaccgaaaa ccggagcctg cctgctcccc ctgagccagc acctgcggga gctgcactac 1021 agccctggcc tccaccccca ccccgccgac catccaaggg agagtgagac ctggcagcca 1081 caactgcagt cccatcctct tgtcagggcc ctttcctgtg tacacgtgac agagtgcctt 1141 ttcgagactg gcagggacga ggacaaatat ggatgaggtg gagagtggga agcaggagcc 1201 cagccagctg cgcctgcgct gcaggagggc gggggctctg gttgtaaaac acacttcctg 1261 ctgcgaaaga cccacatgct acaagacggg caaaataaag tgacagatga ccaccctgca

PD-1 1 agtttccctt ccgctcacct ccgcctgagc agtggagaag gcggcactct ggtggggctg

(SEQ ID NO: 9) 61 ctccaggcat gcagatccca caggcgccct ggccagtcgt ctgggcggtg ctacaactgg

121 gctggcggcc aggatggttc ttagactccc cagacaggcc ctggaacccc cccaccttct 181 ccccagccct gctcgtggtg accgaagggg acaacgccac cttcacctgc agcttctcca 241 acacatcgga gagcttcgtg ctaaactggt accgcatgag ccccagcaac cagacggaca 301 agctggccgc cttccccgag gaccgcagcc agcccggcca ggactgccgc ttccgtgtca 361 cacaactgcc caacgggcgt gacttccaca tgagcgtggt cagggcccgg cgcaatgaca 421 gcggcaccta cctctgtggg gccatctccc tggcccccaa ggcgcagatc aaagagagcc 481 tgcgggcaga gctcagggtg acagagagaa gggcagaagt gcccacagcc caccccagcc 541 cctcacccag gccagccggc cagttccaaa ccctggtggt tggtgtcgtg ggcggcctgc 601 tgggcagcct ggtgctgcta gtctgggtcc tggccgtcat ctgctcccgg gccgcacgag

661 ggacaatagg agccaggcgc accggccagc ccctgaagga ggacccctca gccgtgcctg

721 tgttctctgt ggactatggg gagctggatt tccagtggcg agagaagacc ccggagcccc

781 ccgtgccctg tgtccctgag cagacggagt atgccaccat tgtctttcct agcggaatgg

841 gcacctcatc ccccgcccgc aggggctcag ctgacggccc tcggagtgcc cagccactga

901 ggcctgagga tggacactgc tcttggcccc tctgaccggc ttccttggcc accagtgttc

961 tgcagaccct ccaccatgag cccgggtcag cgcatttcct caggagaagc aggcagggtg

1021 caggccattg caggccgtcc aggggctgag ctgcctgggg gcgaccgggg ctccagcctg

1081 cacctgcacc aggcacagcc ccaccacagg actcatgtct caatgcccac agtgagccca

1141 ggcagcaggt gtcaccgtcc cctacaggga gggccagatg cagtcactgc ttcaggtcct

1201 gccagcacag agctgcctgc gtccagctcc ctgaatctct gctgctgctg ctgctgctgc

1261 tgctgctgcc tgcggcccgg ggctgaaggc gccgtggccc tgcctgacgc cccggagcct

1321 cctgcctgaa cttgggggct ggttggagat ggccttggag cagccaaggt gcccctggca

1381 gtggcatccc gaaacgccct ggacgcaggg cccaagactg ggcacaggag tgggaggtac

1441 atggggctgg ggactcccca ggagttatct gctccctgca ggcctagaga agtttcaggg

1501 aaggtcagaa gagctcctgg ctgtggtggg cagggcagga aacccctcca cctttacaca

1561 tgcccaggca gcacctcagg ccctttgtgg ggcagggaag ctgaggcagt aagcgggcag

1621 gcagagctgg aggcctttca ggcccagcca gcactctggc ctcctgccgc cgcattccac

1681 cccagcccct cacaccactc gggagaggga catcctacgg tcccaaggtc aggagggcag

1741 ggctggggtt gactcaggcc cctcccagct gtggccacct gggtgttggg agggcagaag

1801 tgcaggcacc tagggccccc catgtgccca ccctgggagc tctccttgga acccattcct

1861 gaaattattt aaaggggttg gccgggctcc caccagggcc tgggtgggaa ggtacaggcg

1921 ttcccccggg gcctagtacc cccgccgtgg cctatccact cctcacatcc acacactgca

1981 cccccactcc tggggcaggg ccaccagcat ccaggcggcc agcaggcacc tgagtggctg

2041 ggacaaggga tcccccttcc ctgtggttct attatattat aattataatt aaatatgaga

2101 gcatgctaag gaaaa

CD137L 1 aaaaagcggc gcgctgtgtc ttcccgcagt ctctcgtcat ggaatacgcc tctgacgctt

(SEQ ID NO: 61 cactggaccc cgaagccccg tggcctcccg cgccccgcgc tcgcgcctgc cgcgtactgc

10) 121 cttgggccct ggtcgcgggg ctgctgctgc tgctgctgct cgctgccgcc tgcgccgtct

181 tcctcgcctg cccctgggcc gtgtccgggg ctcgcgcctc gcccggctcc gcggccagcc

241 cgagactccg cgagggtccc gagctttcgc ccgacgatcc cgccggcctc ttggacctgc

301 ggcagggcat gtttgcgcag ctggtggccc aaaatgttct gctgatcgat gggcccctga

361 gctggtacag tgacccaggc ctggcaggcg tgtccctgac ggggggcctg agctacaaag

421 aggacacgaa ggagctggtg gtggccaagg ctggagtcta ctatgtcttc tttcaactag

481 agctgcggcg cgtggtggcc ggcgagggct caggctccgt ttcacttgcg ctgcacctgc

541 agccactgcg ctctgctgct ggggccgccg ccctggcttt gaccgtggac ctgccacccg

601 cctcctccga ggctcggaac tcggccttcg gtttccaggg ccgcttgctg cacctgagtg

661 ccggccagcg cctgggcgtc catcttcaca ctgaggccag ggcacgccat gcctggcagc

721 ttacccaggg cgccacagtc ttgggactct tccgggtgac ccccgaaatc ccagccggac

781 tcccttcacc gaggtcggaa taacgtccag cctgggtgca gcccacctgg acagagtccg

841 aatcctactc catccttcat ggagacccct ggtgctgggt ccctgctgct ttctctacct

901 caaggggctt ggcaggggtc cctgctgctg acctcccctt gaggaccctc ctcacccact

961 ccttccccaa gttggacctt gatatttatt ctgagcctga gctcagataa tatattatat

1021 atattatata tatatatata tttctattta aagaggatcc tgagtttgtg aatggacttt

1081 tttagaggag ttgttttggg gggggggggg tcttcgacat tgccgaggct ggtcttgaac

1141 tcctggactt agacgatcct cctgcctcag cctcccaagc aactgggatt catcctttct

1201 attaattcat tgtacttatt tgcttatttg tgtgtattga gcatctgtaa tgtgccagca

1261 ttgtgcccag gctagggggc tatagaaaca tctagaaata gactgaaaga aaatctgagt

1321 tatggtaata cgtgaggaat ttaaagactc atccccagcc tccacctcct gtgtgatact

1381 tgggggctag cttttttctt tctttctttt ttttgagatg gtcttgttct gtcaaccagg

1441 ctagaatgca gcggtgcaat catgagtcaa tgcagcctcc agcctcgacc tcccgaggct

1501 caggtgatcc tcccatctca gcctctcgag tagctgggac cacagttgtg tgccaccaca

1561 cttggctaac tttttaattt ttttgcggag acggtattgc tatgttgcca aggttgttta

1621 catgccagta caatttataa taaacactca tttttcctcc ctctgaaaaa aaaaaaaaaa

TIM-3 1 atgttttcac atcttccctt tgactgtgtc ctgctgctgc tgctgctact acttacaagg

(SEQ ID NO: 61 tcctcagaag tggaatacag agcggaggtc ggtcagaatg cctatctgcc ctgcttctac

11) 121 accccagccg ccccagggaa cctcgtgccc gtctgctggg gcaaaggagc ctgtcctgtg

181 tttgaatgtg gcaacgtggt gctcaggact gatgaaaggg atgtgaatta ttggacatcc

241 agatactggc taaatgggga tttccgcaaa ggagatgtgt ccctgaccat agagaatgtg

301 actctagcag acagtgggat ctactgctgc cggatccaaa tcccaggcat aatgaatgat 361 gaaaaattta acctgaagtt ggtcatcaaa ccagccaagg tcacccctgc accgactctg

421 cagagagact tcactgcagc ctttccaagg atgcttacca ccaggggaca tggcccagca 481 gagacacaga cactggggag cctccctgat ataaatctaa cacaaatatc cacattggcc 541 aatgagttac gggactctag attggccaat gacttacggg actctggagc aaccatcaga 601 ataggcatct acatcggagc agggatctgt gctgggctgg ctctggctct tatcttcggc 661 gctttaattt tcaaatggta ttctcatagc aaagagaaga tacagaattt aagcctcatc 721 tctttggcca acctccctcc ctcaggattg gcaaatgcag tagcagaggg aattcgctca 781 gaagaaaaca tctataccat tgaagagaac gtatatgaag tggaggagcc caatgagtat 841 tattgctatg tcagcagcag gcagcaaccc tcacaacctt tgggttgtcg ctttgcaatg 901 ccatag

OX40L 1 ccgcaaggaa aacccagact ctggcgacag cagagacgag gatgtgcgtg ggggctcggc (SEQ ID NO: 61 ggctgggccg cgggccgtgt gcggctctgc tcctcctggg cctggggctg agcaccgtga 12) 121 cggggctcca ctgtgtcggg gacacctacc ccagcaacga ccggtgctgc cacgagtgca

181 ggccaggcaa cgggatggtg agccgctgca gccgctccca gaacacggtg tgccgtccgt 241 gcgggccggg cttctacaac gacgtggtca gctccaagcc gtgcaagccc tgcacgtggt 301 gtaacctcag aagtgggagt gagcggaagc agctgtgcac ggccacacag gacacagtct 361 gccgctgccg ggcgggcacc cagcccctgg acagctacaa gcctggagtt gactgtgccc 421 cctgccctcc agggcacttc tccccaggcg acaaccaggc ctgcaagccc tggaccaact 481 gcaccttggc tgggaagcac accctgcagc cggccagcaa tagctcggac gcaatctgtg 541 aggacaggga ccccccagcc acgcagcccc aggagaccca gggccccccg gccaggccca 601 tcactgtcca gcccactgaa gcctggccca gaacctcaca gggaccctcc acccggcccg 661 tggaggtccc cgggggccgt gcggttgccg ccatcctggg cctgggcctg gtgctggggc 721 tgctgggccc cctggccatc ctgctggccc tgtacctgct ccggagggac cagaggctgc 781 cccccgatgc ccacaagccc cctgggggag gcagtttccg gacccccatc caagaggagc 841 aggccgacgc ccactccacc ctggccaaga tctgacctgg gcccaccaag gtggacgctg 901 ggccccgcca ggctggagcc cggagggtct gctgggcgag cagggcaggt gcaggccgcc 961 tgccccgcca cgctcctggg ccaactctgc accgttctag gtgccgatgg ctgcctccgg 1021 ctctctgctt acgtatgcca tgcatacctc ctgccccgcg ggaccacaat aaaaaccttg 1081 gcagacggga gtctccgacc ggcaaaaaaa aaaaaaaaaa

CTLA4 1 cttctgtgtg tgcacatgtg taatacatat ctgggatcaa agctatctat ataaagtcct

(SEQ ID NO: 61 tgattctgtg tgggttcaaa cacatttcaa agcttcagga tcctgaaagg ttttgctcta 13) 121 cttcctgaag acctgaacac cgctcccata aagccatggc ttgccttgga tttcagcggc

181 acaaggctca gctgaacctg gctaccagga cctggccctg cactctcctg ttttttcttc 241 tcttcatccc tgtcttctgc aaagcaatgc acgtggccca gcctgctgtg gtactggcca 301 gcagccgagg catcgccagc tttgtgtgtg agtatgcatc tccaggcaaa gccactgagg 361 tccgggtgac agtgcttcgg caggctgaca gccaggtgac tgaagtctgt gcggcaacct 421 acatgatggg gaatgagttg accttcctag atgattccat ctgcacgggc acctccagtg 481 gaaatcaagt gaacctcact atccaaggac tgagggccat ggacacggga ctctacatct 541 gcaaggtgga gctcatgtac ccaccgccat actacctggg cataggcaac ggaacccaga 601 tttatgtaat tgatccagaa ccgtgcccag attctgactt cctcctctgg atccttgcag 661 cagttagttc ggggttgttt ttttatagct ttctcctcac agctgtttct ttgagcaaaa

721 tgctaaagaa aagaagccct cttacaacag gggtctatgt gaaaatgccc ccaacagagc 781 cagaatgtga aaagcaattt cagccttatt ttattcccat caattgagaa accattatga 841 agaagagagt ccatatttca atttccaaga gctgaggcaa ttctaacttt tttgctatcc 901 agctattttt atttgtttgt gcatttgggg ggaattcatc tctctttaat ataaagttgg

961 atgcggaacc caaattacgt gtactacaat ttaaagcaaa ggagtagaaa gacagagctg 1021 ggatgtttct gtcacatcag ctccactttc agtgaaagca tcacttggga ttaatatggg 1081 gatgcagcat tatgatgtgg gtcaaggaat taagttaggg aatggcacag cccaaagaag 1141 gaaaaggcag ggagcgaggg agaagactat attgtacaca ccttatattt acgtatgaga 1201 cgtttatagc cgaaatgatc ttttcaagtt aaattttatg ccttttattt cttaaacaaa

1261 tgtatgatta catcaaggct tcaaaaatac tcacatggct atgttttagc cagtgatgct 1321 aaaggttgta ttgcatatat acatatatat atatatatat atatatatat atatatatat

1381 atatatatat tttaatttga tagtattgtg catagagcca cgtatgtttt tgtgtatttg

1441 ttaatggttt gaatataaac actatatggc agtgtctttc caccttgggt cccagggaag 1501 ttttgtggag gagctcagga cactaataca ccaggtagaa cacaaggtca tttgctaact 1561 agcttggaaa ctggatgagg tcatagcagt gcttgattgc gtggaattgt gctgagttgg 1621 tgttgacatg tgctttgggg cttttacacc agttcctttc aatggtttgc aaggaagcca 1681 cagctggtgg tatctgagtt gacttgacag aacactgtct tgaagacaat ggcttactcc 1741 aggagaccca caggtatgac cttctaggaa gctccagttc gatgggccca attcttacaa 1801 acatgtggtt aatgccatgg acagaagaag gcagcaggtg gcagaatggg gtgcatgaag 1861 gtttctgaaa attaacactg cttgtgtttt taactcaata ttttccatga aaatgcaaca

1921 acatgtataa tatttttaat taaataaaaa tctgtggtgg tcgttttaaa aaaaaaaaaa

1981 aaaaaaaaaa aaaaaaaaaa aaaaaaaaaa aaaaaaaaaa aaaaa

VISTA i g ggggcgggt gcctggagca cggcgctggg gccgcccgca gcgctcactc gctcgcactc

(SEQ ID NO: 61 agtcgcggga ggcttccccg cgccggccgc gtcccgcccg ctccccggca ccagaagttc

14) 121 ctctgcgcgt ccgacggcga catgggcgtc cccacggccc tggaggccgg cagctggcgc

181 tggggatccc tgctcttcgc tctcttcctg gctgcgtccc taggtccggt ggcagccttc

241 aaggtcgcca cgccgtattc cctgtatgtc tgtcccgagg ggcagaacgt caccctcacc

301 tgcaggctct tgggccctgt ggacaaaggg cacgatgtga ccttctacaa gacgtggtac

361 ^gcagctcga ggggcgaggt gcagacctgc tcagagcgcc ggcccatccg caacctcacg

421 ttccaggacc ttcacctgca ccatggaggc caccaggctg ccaacaccag ccacgacctg

481 gctcagcgcc acgggctgga gtcggcctcc gaccaccatg gcaacttctc catcaccatg

541 Egcaacctga ccctgctgga tagcggcctc tactgctgcc tggtggtgga gatcaggcac

601 caccactcgg agcacagggt ccatggtgcc atggagctgc aggtgcagac aggcaaagat

661 gcaccatcca actgtgtggt gtacccatcc tcctcccagg atagtgaaaa catcacggct

721 gcagccctgg ctacgggtgc ctgcatcgta ggaatcctct gcctccccct catcctgctc

781 ctggtctaca agcaaaggca ggcagcctcc aaccgccgtg cccaggagct ggtgcggatg

841 gacagcaaca ttcaagggat tgaaaacccc ggctttgaag cctcaccacc tgcccagggg

901 atacccgagg ccaaagtcag gcaccccctg tcctatgtgg cccagcggca gccttctgag

961 tctgggcggc atctgctttc ggagcccagc acccccctgt ctcctccagg ccccggagac

1021 gtcttcttcc catccctgga ccctgtccct gactctccaa actttgaggt catctagccc

1081 agctggggga cagtgggctg ttgtggctgg gtctggggca ggtgcatttg agccagggct

1141 ggctctgtga gtggcctcct tggcctcggc cctggttccc tccctcctgc tctgggctca

1201 gatactgtga catcccagaa gcccagcccc tcaacccctc tggatgctac atggggatgc

1261 tggacggctc agcccctgtt ccaaggattt tggggtgctg agattctccc ctagagacct

1321 gaaattcacc agctacagat gccaaatgac ttacatctta agaagtctca gaacgtccag

1381 cccttcagca gctctcgttc tgagacatga gccttgggat gtggcagcat cagtgggaca

1441 agatggacac tgggccaccc tcccaggcac cagacacagg gcacggtgga gagacttctc

1501 ccccgtggcc gccttggctc ccccgttttg cccgaggctg ctcttctgtc agacttcctc

1561 tttgtaccac agtggctctg gggccaggcc tgcctgccca ctggccatcg ccaccttccc

1621 cagctgcctc ctaccagcag tttctctgaa gatctgtcaa caggttaagt caatctgggg

1681 cttccactgc ctgcattcca gtccccagag cttggtggtc ccgaaacggg aagtacatat

1741 tggggcatgg tggcctccgt gagcaaatgg tgtcttgggc aatctgaggc caggacagat

1801 gttgccccac ccactggaga tggtgctgag ggaggtgggt ggggccttct gggaaggtga

1861 gtggagaggg gcacctgccc cccgccctcc ccatccccta ctcccactgc tcagcgcggg

1921 ccattgcaag ggtgccacac aatgtcttgt ccaccctggg acacttctga gtatgaagcg

1981 ggatgctatt aaaaactaca tggggaaaca ggtgcaaacc ctggagatgg attgtaagag

2041 ccagtttaaa tctgcactct gctgctcctc ccccaccccc accttccact ccatacaatc

2101 tgggcctggt ggagtcttcg cttcagagcc attcggccag gtgcgggtga tgttcccatc

2161 tcctgcttgt gggcatgccc tggctttgtt tttatacaca taggcaaggt gagtcctctg

2221 tggaattgtg attgaaggat tttaaagcag gggaggagag tagggggcat ctctgtacac

2281 tctgggggta aaacagggaa ggcagtgcct gagcatgggg acaggtgagg tggggctggg

2341 cagaccccct gtagcgttta gcaggatggg ggccccaggt actgtggaga gcatagtcca

2401 gcctgggcat ttgtctccta gcagcctaca ctggctctgc tgagctgggc ctgggtgctg

2461 aaagccagga tttggggcta ggcgggaaga tgttcgccca attgcttggg gggttggggg

2521 gatggaaaag gggagcacct ctaggctgcc tggcagcagt gagccctggg cctgtggcta

2581 cagccaggga accccacctg gacacatggc cctgcttcta agccccccag ttaggcccaa

2641 aggaatggtc cactgagggc ctcctgctct gcctgggctg ggccaggggc tttgaggaga

2701 gggtaaacat aggcccggag atggggctga cacctcgagt ggccagaata tgcccaaacc

2761 ccggcttctc ccttgtccct aggcagaggg gggtcccttc ttttgttccc tctggtcacc

2821 acaatgcttg atgccagctg ccataggaag agggtgctgg ctggccatgg tggcacacac

2881 ctgtcctccc agcactttgc agggctgagg tggaaggacc gcttaagccc aggtgttcaa

2941 ggctgctgtg agctgtgttc gagccactac actccagcct ggggacggag caaaactttg

3001 cctcaaaaca aattttaaaa agaaagaaag aaggaaagag ggtatgtttt tcacaattca

3061 tgggggcctg catggcagga gtggggacag gacacctgct gttcctggag tcgaaggaca

3121 agcccacagc ccagattccg gttctcccaa ctcaggaaga gcatgccctg ccctctgggg

3181 aggctggcct ggccccagcc ctcagctgct gaccttgagg cagagacaac ttctaagaat

3241 ttggctgcca gaccccaggc ctggctgctg ctgtgtggag agggaggcgg cccgcagcag

3301 aacagccacc gcacttcctc ctcagcttcc tctggtgcgg ccctgccctc tcttctctgg

3361 acccttttac aactgaacgc atctgggctt cgtggtttcc tgttttcagc gaaatttact 3421 ctgagctccc agttccatct tcatccatgg ccacaggccc tgcctacaac gcactaggga

3481 cgtccctccc tgctgctgct ggggaggggc aggctgctgg agccgccctc tgagttgccc 3541 gggatggtag tgcctctgat gccagccctg gtggctgtgg gctggggtgc atgggagagc 3601 tgggtgcgag aacatggcgc ctccaggggg cgggaggagc actaggggct ggggcaggag 3661 gctcctggag cgctggattc gtggcacagt ctgaggccct gagagggaaa tccatgcttt 3721 taagaactaa ttcattgtta ggagatcaat caggaattag gggccatctt acctatctcc 3781 tgacattcac agtttaatag agacttcctg cctttattcc ctcccaggga gaggctgaag 3841 gaatggaatt gaaagcacca tttggagggt tttgctgaca cagcggggac tgctcagcac 3901 tccctaaaaa cacaccatgg aggccactgg tgactgctgg tgggcaggct ggccctgcct 3961 gggggagtcc gtggcgatgg gcgctggggt ggaggtgcag gagccccagg acctgctttt 4021 caaaagactt ctgcctgacc agagctccca ctacatgcag tggcccaggg cagaggggct 4081 gatacatggc ctttttcagg gggtgctcct cgcggggtgg acttgggagt gtgcagtggg 4141 acagggggct gcaggggtcc tgccaccacc gagcaccaac ttggcccctg gggtcctgcc 4201 tcatgaatga ggccttcccc agggctggcc tgactgtgct gggggctggg ttaacgtttt 4261 ctcagggaac cacaatgcac gaaagaggaa ctggggttgc taaccaggat gctgggaaca 4321 aaggcctctt gaagcccagc cacagcccag ctgagcatga ggcccagccc atagacggca 4381 caggccacct ggcccattcc ctgggcattc cctgctttgc attgctgctt ctcttcaccc 4441 catggaggct atgtcaccct aactatcctg gaatgtgttg agagggattc tgaatgatca 4501 atatagcttg gtgagacagt gccgagatag atagccatgt ctgccttggg cacgggagag 4561 ggaagtggca gcatgcatgc tgtttcttgg ccttttctgt tagaatactt ggtgctttcc 4621 aacacacttt cacatgtgtt gtaacttgtt tgatccaccc ccttccctga aaatcctggg 4681 aggttttatt gctgccattt aacacagagg gcaatagagg ttctgaaagg tctgtgtctt

4741 gtcaaaacaa gtaaacggtg gaactacgac taaa

CD200 1 gtcagtttcc ccagcggtca cctttgaaaa gggaaaaatg tctgaaaata gacaaagctg (SEQ ID NO: 61 aatataaaca tcatttaatt ccccccacac agacagcctc cgctcctgtg agggcgtggg 15) 121 gaaaacggag tgggagaagg gggctagcga ggaggaagag gcgggaggtg cggcaggggc

181 acaggtgacg ctcctcccgc ctgcctagca gagctccagg cgcacatccg cagtcagcca

241 cctcgcgcgc gcctccagga gcaaggatgg agaggctggt gatcaggatg cccttctctc

301 atctgtctac ctacagcctg gtttgggtca tggcagcagt ggtgctgtgc acagcacaag

361 tgcaagtggt gacccaggat gaaagagagc agctgtacac acctgcttcc ttaaaatgct

421 ctctgcaaaa tgcccaggaa gccctcattg tgacatggca gaaaaagaaa gctgtaagcc

481 cagaaaacat ggtcaccttc agcgagaacc atggggtggt gatccagcct gcctataagg

541 acaagataaa cattacccag ctgggactcc aaaactcaac catcaccttc tggaatatca

601 ccctggagga tgaagggtgt tacatgtgtc tcttcaatac ctttggtttt gggaagatct

661 caggaacggc ctgcctcacc gtctatgtac agcccatagt atcccttcac tacaaattct

721 ctgaagacca cctaaatatc acttgctctg ccactgcccg cccagccccc atggtcttct

781 ggaaggtccc tcggtcaggg attgaaaata gtacagtgac tctgtctcac ccaaatggga

841 ccacgtctgt taccagcatc ctccatatca aagaccctaa gaatcaggtg gggaaggagg

901 tgatctgcca ggtgctgcac ctggggactg tgaccgactt taagcaaacc gtcaacaaag

961 gctattggtt ttcagttccg ctattgctaa gcattgtttc cctggtaatt cttctcgtcc

1021 taatctcaat cttactgtac tggaaacgtc accggaatca ggaccgagag ccctaaataa

1081 gtcacacagc accctgaaag tgattccctg gtctacttga atttgacaca agagaaaagc

1141 aggaggaaaa ggggccattc tccaaaggac ctgaaagagc aaaagaggtg ggagcgaaag

1201 ccttaaggat cccacgactt tttactgcca tctgagctac tcagtgtttg aatcccaaga

1261 ggaagtcagt ttacctctca ggtctgttgt aggacttgat tttgtaaagc aatgccatgt

1321 tatgtggttg aaagggcact ggacttagtt agtatcagga gcactgagct cacagactga

1381 cttgggctcc tactggtggg gacctctgtt agtcacttta cctcatccaa agtataaagg

1441 aattggacca aataatttac cacatagctc taaaacttaa tttaaaatgt aattccagaa

1501 aaaaaaaggg aataagcaaa gggggaagaa ttgaaagaga gagagaagaa agaatacaga

1561 gagcttacct tttgcctttc tgttgatgtt acatctcttc ttcctatgtt cttaggtcta

1621 tgagtctgtt tccccatcat ttggtatcta gtccagttcc tgcttactgc tttgctaata

1681 gctggccttg ctagaatcct tggtttcact gctgttcttc atgtgcttct atgagattta

1741 ctccaacaca aataggactg aatttattgt gaagtaacat tggcaatctt aacttattca

1801 tttaacttat ttttatagct agataaatat tgttagtctt agacaatagc tcacattttt

1861 tgagaagcat gccctccctg tccatttgtc ttataacatg acccagccct attttacgtc

1921 attctaaatt cagcctcata taatgaaaat acattatgaa aacagatgtt taggagattt

1981 cctgtatagc agtcagccaa ttcatatgct ttgtctctgc tggcttcttt ttccatgcgt

2041 taacttttcc caatagcaga ggaggcaaat atgagcatac aatccctttg ttctaaagat

2101 attgttccag ctagtggaat gatgttgaat ctttaataac cataattagt tgctttttca

2161 gtatcttctg ctttgtctgt gtctatccag tggcctagga attaaagtgt aagttgtttt 2221 cgctgttaaa ttggatattt atatatatat atagcaagat tttcatgtgt tatttaattc

2281 tgtattgttt cttatatttg tagtaaaata ttgaacaatt aaaagtgttg actccaaaaa

2341 aaaaaaaa

[0034] As used herein the terms "disease or disorder" is any one of a group of ailments capable of causing an negative health in a subject by: (i) expression of one or a plurality of mutated nucleic acid sequences in one or a plurality of amino acids; or (ii) aberrant expression of one or a plurality of nucleic acid sequences in one or a plurality of amino acids, in each case, in an amount that causes an abnormal biological affect that negatively affects the health of the subject. In some embodiments, the disease or disorder is chosen from: cancer of the adrenal gland, bladder, bone, bone marrow, brain, spine, breast, cervix, gall bladder, ganglia, gastrointestinal tract, stomach, colon, heart, kidney, liver, lung, muscle, ovary, pancreas, parathyroid, penis, prostate, salivary glands, skin, spleen, testis, thymus, thyroid, or uterus. In some embodiments, a disease or disorder is a hyperproliferative disease. The term hyperproliferative disease means a cancer chosen from: lung cancer, bone cancer, CMML, pancreatic cancer, skin cancer, cancer of the head and neck, cutaneous or intraocular melanoma, uterine cancer, ovarian cancer, rectal cancer, cancer of the anal region, stomach cancer, colon cancer, breast cancer, testicular, gynecologic tumors (e.g., uterine sarcomas, carcinoma of the fallopian tubes, carcinoma of the endometrium, carcinoma of the cervix, carcinoma of the vagina or carcinoma of the vulva), Hodgkin's disease, cancer of the esophagus, cancer of the small intestine, cancer of the endocrine system (e.g., cancer of the thyroid, parathyroid or adrenal glands), sarcomas of soft tissues, cancer of the urethra, cancer of the penis, prostate cancer, chronic or acute leukemia, solid tumors of childhood, lymphocytic lymphomas, cancer of the bladder, cancer of the kidney or ureter (e.g., renal cell carcinoma, carcinoma of the renal pelvis), or neoplasms of the central nervous system (e.g., primary CNS lymphoma, spinal axis tumors, brain stem gliomas or pituitary adenomas).

[0035] As used herein the terms "electronic medium" mean any physical storage employing electronic technology for access, including a hard disk, ROM, EEPROM, RAM, flash memory, nonvolatile memory, or any substantially and functionally equivalent medium. In some

embodiments, the software storage may be co-located with the processor implementing an embodiment of the invention, or at least a portion of the software storage may be remotely located but accessible when needed.

[0036] As used herein, the terms "information associated with the disease or disorder" means any information related to a disease or disorder necessary to perform the method described herein or to run the software identified herein. In some embodiments, the information associated with a disease or disorder is any information from a subject that can be used or is used as a parameter or variable in the input of any analytical function performed in the course of performing any method disclosed herein. In some embodiments, the information associated with the disease or disorder is selected from: DNA or RNA expression levels of a subject or population of subjects, amino acid expression levels of a subject or population of subjects, whether or not the subject or population is taking a therapy for a condition, the age of a subject or population of subjects, the gender of a subject or population of subjects; or whether and, if so, how much or how long a subject or population of subjects has been diagnosed with a cancer disclosed herein, and/or exposed to an environmental condition, drug or biologic.

[0037] As used herein, "inhibitors" or "antagonists" of a given protein refer to modulatory molecules or compounds that, e.g., bind to, partially or totally block activity, decrease, prevent, delay activation, inactivate, desensitize, or down regulate the activity or expression of the given protein, or downstream molecules regulated by such a protein. Inhibitors can include siRNA or antisense RNA, genetically modified versions of the protein, e.g., versions with altered activity, as well as naturally occurring and synthetic antagonists, antibodies, small chemical molecules and the like. Assays for identifying other inhibitors can be performed in vitro or in vivo, e.g., in cells, or cell membranes, by applying test inhibitor compounds, and then determining the functional effects on activity.

[0038] The term "nucleic acid" refers to a molecule comprising two or more linked nucleotides. "Nucleic acid" and "nucleic acid molecule" are used interchangeably and refer to oligoribonucleotides as well as oligodeoxyribonucleotides. The terms also include polynucleosides (i.e., a polynucleotide minus a phosphate) and any other organic base containing nucleic acid. The organic bases include adenine, uracil, guanine, thymine, cytosine and inosine. The nucleic acids may be single or double stranded. The nucleic acid may be naturally or non-naturally occurring. Nucleic acids can be obtained from natural sources, or can be synthesized using a nucleic acid synthesizer (i.e., synthetic). Isolation of nucleic acids are routinely performed in the art and suitable methods can be found in standard molecular biology textbooks. (See, for example, Maniatis' Handbook of Molecular Biology.) The nucleic acid may be DNA or RNA, such as genomic DNA, mitochondrial DNA, mRNA, cDNA, rRNA, miRNA, PNA or LNA, or a combination thereof, as described herein. In some embodiments, the term nucleic acid sequence is used to refer to expression of genes with all or part of their regulatory sequences operably linked to the expressible components of the gene. In some embodiments, the expression of genes is analyzed for genetic interactions. In other embodiments, genetic interactions are analyzed by identifying pairs of a first gene and a second gene whose expression or activity contributes to the modulation of the lethality or likelihood of a subject from which the information associated with a disease or disorder is obtained. In some embodiments, the nucleic acid pair (comprising a first and second nucleic acid) is a pair of microRNAs, shRNAs, amino acids or nucleic acid sequences defined with presence of only partial regulatory sequences operably linked to the expressible components of a gene.

[0039] Some aspects of this invention relate to the use of nucleic acid derivatives or synthetic sequences. The use of certain nucleic acid derivatives or synthetic sequences may enable complementarity as between natural expression products (such as mRNA) and the synthetic sequences to block protein translation of products for validation of software analysis and corroboration with biological assays. As used herein, a nucleic acid derivative is a non-naturally occurring nucleic acid or a unit thereof. Nucleic acid derivatives may contain non-naturally occurring elements such as non-naturally occurring nucleotides and non- naturally occurring backbone linkages. Nucleic acid derivatives according to some aspects of this invention may contain backbone modifications such as but not limited to phosphorothioate linkages, phosphodiester modified nucleic acids, combinations of phosphodiester and phosphorothioate nucleic acid, methylphosphonate, alkylphosphonates, phosphate esters, alkylphosphonothioates, phosphoramidates, carbamates, carbonates, phosphate triesters, acetamidates, carboxymethyl esters, methylphosphorothioate, phosphorodithioate, p-ethoxy, and combinations thereof. The backbone composition of the nucleic acids may be homogeneous or heterogeneous. Nucleic acid derivatives according to some aspects of this invention may contain substitutions or modifications in the sugars and/or bases. For example, some nucleic acid derivatives may include nucleic acids having backbone sugars which are covalently attached to low molecular weight organic groups other than a hydroxyl group at the 3' position and other than a phosphate group at the 5' position (e.g., an 2'-0-alkylated ribose group). Nucleic acid derivatives may include non-ribose sugars such as arabinose. Nucleic acid derivatives may contain substituted purines and pyrimidines such as C-5 propyne modified bases, 5-methylcytosine, 2-aminopurine, 2-amino-6- chloropurine, 2,6-diaminopurine, hypoxanthine, 2-thiouracil and pseudoisocytosine. In some embodiments, a nucleic acid may comprise a peptide nucleic acid (PNA), a locked nucleic acid (LNA), DNA, RNA, or a co-nucleic acids of the above such as DNA-LNA co-nucleic acid. [0040] As used herein, the term "probability score" refers to a quantitative value given to the output of any one or series of algorithms that are disclosed herein. In some embodiments, the probability score is determined by application of one or plurality of algorithm disclosed herein by: setting, by the at least one processor, a predetermined value, stored in the memory, that corresponds to a threshold value above which the first pair of nucleic acid sequence is correlated to an interaction event, the ineffectiveness or effectiveness of a therapy, the resistance of a therapy, and/or the prognosis of the subject or population of subjects suffering from a disease or disorder; calculating, by the at least one processor, the probability score, wherein calculating the probability score comprises: (i) analyzing information associated with a disease or disorder of the subject or the population of subjects; and (ii) conducting one or a plurality of statistical tests from the information associated with a disease or disorder; and (iii) assigning a probability score related to an interaction event, the ineffectiveness or effectiveness of a therapy, the resistance of a therapy, and/or the prognosis of the subject or population of subjects suffering from cancer based upon a comparison of outcomes from the operation of statistical tests and the threshold value. In some embodiments, the threshold value is a calculated area under the curve (AUC) of a Receiver Operating Characteristic (ROC) curve. In some embodiments, the threashold value is an AUC calculation that is at least about 0.5, 0.6, 0.7, 0.8, 0.9, 1.0 or greater in its ROC curve.

In some embodiments, the AUC calculation contributes to the probability score associated with any of the steps of any of the methods disclosed herein and is from about 0.5 to about 1.5. In some embodiments, the AUC calculation contributes to the probability score associated with any of the steps of any of the methods disclosed herein and is from about 0.55 to about 1.5. In some embodiments, the AUC calculation contributes to the probability score associated with any of the steps of any of the methods disclosed herein and is from about 0.6 to about 1.5. In some embodiments, the AUC calculation contributes to the probability score associated with any of the steps of any of the methods disclosed herein and is from about 0.65 to about 1.5. In some embodiments, the AUC calculation contributes to the probability score associated with any of the steps of any of the methods disclosed herein and is from about 0.70 to about 1.5. In some embodiments, the AUC calculation contributes to the probability score associated with any of the steps of any of the methods disclosed herein and is from about 0.75 to about 1.5. In some embodiments, the AUC calculation contributes to the probability score associated with any of the steps of any of the methods disclosed herein and is from about 0.80 to about 1.5. In some embodiments, the AUC calculation contributes to the probability score associated with any of the steps of any of the methods disclosed herein and is from about 0.85 to about 1.5. In some embodiments, the AUC calculation contributes to the probability score associated with any of the steps of any of the methods disclosed herein and is from about 0.90 to about 1.5. In some embodiments, the AUC calculation contributes to the probability score associated with any of the steps of any of the methods disclosed herein and is from about 0.7 to about 0.8.

[0041] As used herein, the terms "peptide," "polypeptide" and "protein" are used interchangeably and refer to two or more amino acids covalently linked by an amide bond or non-amide equivalent. The peptides of the disclosure can be of any length. For example, the peptides can have from about two to about 100 or more residues, such as, 5 to 12, 12 to 15, 15 to 18, 18 to 25, 25 to 50, 50 to 75, 75 to 100, or more in length. Preferably, peptides are from about 2 to about 18 residues in length. The peptides of the disclosure also include 1- and d-isomers, and combinations of 1- and d-isomers. The peptides can include modifications typically associated with posttranslational processing of proteins, for example, cyclization (e.g., disulfide or amide bond), phosphorylation, glycosylation, carboxylation, ubiquitination, myristylation, or lipidation.

[0042] As used herein, the term "prognosing" means determining the probable course and/or clinical outcome of a disease.

[0043] Any probes may be used in concert with any of the devices, computer program product or methods disclosed herein. As used herein, the term "probe" refers to any molecule that may bind or associate, indirectly or directly, covalently or non-covalently, to any of the nucleic acid sequences disclosed herein or any immune checkpoint proteins or variants thereof encoded by the nucleic acid sequences disclosed herein and whose association or binding is detectable using the methods disclosed herein. In some embodiments, the probe is a fluorogenic, fluorescent, or chemiluminescent probe, an antibody, or an absorbance-based probe. In some embodiments, an absorbance-based probe, for example the chromophore pNA (para-nitroanaline), may be used as a probe for detection and/or quantification of a protease disclosed herein. In some embodiments, the probe comprises an amino acid sequence that naturally binds to any of the nucleic acid sequences disclosed herein or any immune checkpoint proteins or variants thereof encoded by the nucleic acid sequences, including those variants that are at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 87%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, or at least about 99% homologous to amino acids encoded by SEQ ID NOs: 1 through 15 below. In the case of the disclosewd nucleic acid sequence, a probe may be a complementary nucleic acid sequence that is comprises at least about 4, 5, 6, 7, 8, 9, or 10 or more nucleic acids complementary to any of the nucleic acid sequences disclosed herein or variants thereof that are at least 70% homologous to any of SEQ ID NO: 1 through 15. A probe may be immobilized, adsorbed, or otherwise non- covalently bound to a solid surface, such that upon exposure to any of the nucleic acid sequences disclosed herein or any immune checkpoint proteins or variants thereof encoded by the nucleic acid sequences for a time period sufficient to perform a detectable reaction. In some embodiments, cleavage of a substrate bound to the any of the nucleic acid sequences disclosed herein or any immune checkpoint proteins or variants thereof encoded by the nucleic acid sequences, non-covalently or covantly, causes a biological change in the nature or chemical availability of one or more probes such that cleavage enables detection of the reaction product. For instance, if the step of detecting comprises use of FRET, cleavage of the substrate disclosed herein causes one of the chromophore to emit a fluorescent light under exposure to a wavelength sufficient to activate such a fluorescent molecule. The intensity, length, or amplitude of a wavelength emitted from fluorescent marker can be measured and is, in some embodiments, proportional to the presence, absence or quantity of any of the nucleic acid sequences disclosed herein or any immune checkpoint proteins or variants thereof encoded by the nucleic acid sequences present in the sample, reaction vessel comprising the sample, thereby the quantity of enzyme can be determined from detection of the intensity of or fluorescence at a known wavelength of light.

[0044] An "activity-based probe," as used herein, refers to a certain embodiment of probe comprising a small molecule that binds to or has affinity for a molecule such as a substrate that binds any of the nucleic acid sequences disclosed herein or any immune checkpoint proteins or variants thereof encoded by the nucleic acid sequences in the presence of such nucleic acid sequences disclosed herein or any immune checkpoint proteins or variants thereof encoded by the nucleic acid sequences, such that its bound or unbound state confers an activity readout detectable by PCR, fluorescence, absorbance or any other detection means. In some embodiments, the activity-based probe covalently or non- covalently binds to any of the nucleic acid sequences disclosed herein or any immune checkpoint proteins or variants thereof encoded by the nucleic acid sequences. In some embodiments, the binding of the activity-based probe modifies the physical or biological activity of the any of the nucleic acid sequences disclosed herein or any immune checkpoint proteins or variants thereof encoded by the nucleic acid sequences. In some embodiments, the activity-based probe can be fluorescent or chemiluminescent. In some embodiments, the activity-based probe has a measurable activity of one value if any of the nucleic acid sequences disclosed herein or any immune checkpoint proteins or variants thereof encoded by the nucleic acid sequences are inactive and another measurable activity if in an activated state.

[0045] As used herein, the terms "fluorogenic" and "fluorescent" probe refer to any molecule (dye, quantum dot, peptide, or fluorescent marker) that emits a known and/or detectable wavelength of light upon exposure to a known wavelength of light. In some embodiments, the substrates or peptides with known cleavage sites recognizable by any of the nucleic acid sequences disclosed herein or any immune checkpoint proteins or variants thereof encoded by the nucleic acid sequences are covalently or non-covalently attached to a fluorogenic probe. In some embodiments, the attachment of the fluorogenic probe to the substrate creates a chimeric molecule capable of a fluorescent emission or emissions upon exposure of the substrate to the immune checkpoint proteins or variants and the known wavelength of light, such that exposure to the immune checkpoint proteins or variants thereof creates a reaction product which is quantifiable in the presence of a fluorimeter. In some embodiments, light from the fluorogenic probe is fully quenched upon exposure to the known wavelength of light before enzymatic cleavage of the substrate and the fluorogenic probe emits a known wavelength of light, the intensity of which is quantifiable by absorbance readings or intensity levels in the presence of a fluorimeter and after enzymatic cleavage of the substrate. In some embodiments, the fluorogenic probe is a coumarin-based dye or rhodamine-based dye with fluorescent emission spectra measureable or quantifiable in the presence of or exposure to a predetermined wavelength of light. In some embodiments, the fluorogenic probe comprises rhodamine. In some embodiments, the fluorogenic probe comprises rhodamine- 100. Coumarin-based fluorogenic probes are known in the art, for example in US Pat Nos. 7,625,758 and 7,863,048, which are herein incorporated by reference in their entireties. In some embodiments, the fluorogenic probes are a component to, covalently bound to, non-covalently bound to, intercalated with one or a plurality of substrates to any of the immune checkpoint proteins or variants disclosed herein. In some embodiments, the fluorogenic probes are chosen from ACC or AMC. In some embodiments, the fluorogenic probe is a fluorescein molecule. In some embodiments, the fluorogenic probe is capable of emitting a resonance wave detectable and/or quantifiable by a fluorimeter after exposure to one or a plurality of immune checkpoint proteins or variants disclosed herein. "Fluorescence microscopy," which uses the fluorescence to generate an image, may be used to detect the presence, absence, or quantity of a fluorescent probe. In some embodiments, fluorescence microscopy comprises measuring fluorescence resonance energy transfer (FRET) within a FRET -based assay.

[0046] A "chemiluminescent probe" refers to any molecule (dye, peptide, or chemiluminescent marker) that emits a known and/or detectable wavelength of light as the result of a chemical reaction. Chemiluminescence differs from fluorescence or phosphorescence in that the electronic excited state is the product of a chemical reaction rather than of the absorption of a photon. Non-limiting examples of chemiluminescent probes are luciferin and aequorin molecules. In some embodiments, a chemiluminescent molecule is covalently or non-covalently attached to a nucleic acid or protein encoded by the nucleic acid disclosed herein, such that the excited electronic state can be quantified to determine directly that an enzyme, such as an immune checkpoint protein, is in a reaction vessel, or, indirectly, by quantifying the amount of reaction product was produced after activation of the probe on the substrate or a portion of the substrtate.

[0047] As used herein, the term "sample" refers to a biological sample obtained or derived from a source of interest, as described herein. In some embodiments, a source of interest comprises an organism, such as an animal or human. In some embodiments, a biological sample comprises biological tissue or fluid. In some embodiments, a biological sample may be or comprise bone marrow; blood; blood cells; ascites; tissue or fine needle biopsy samples; cell-containing body fluids; free floating nucleic acids; sputum; saliva; urine; cerebrospinal fluid, peritoneal fluid; pleural fluid; feces; lymph; gynecological fluids; skin swabs; vaginal swabs; oral swabs; nasal swabs; washings or lavages such as a ductal lavages or broncheoalveolar lavages; aspirates; scrapings; bone marrow specimens; tissue biopsy specimens; surgical specimens; feces, other body fluids, secretions, and/or excretions; and/or cells therefrom, etc. In some embodiments, a biological sample is or comprises bodily fluid. In some embodiments, a sample is a "primary sample" obtained directly from a source of interest by any appropriate means. For example, in some embodiments, a primary biological sample is obtained by methods selected from the group consisting of biopsy (e.g., fine needle aspiration or tissue biopsy), surgery, collection of body fluid (e.g., blood, lymph, feces etc.), etc. In some embodiments, as will be clear from context, the term "sample" refers to a preparation that is obtained by processing (e.g., by removing one or more components of and/or by adding one or more agents to) a primary sample. For example, filtering using a semi-permeable membrane. Such a "processed sample" may comprise, for example nucleic acids or proteins extracted from a sample or obtained by subjecting a primary sample to techniques such as amplification or reverse transcription of mRNA, isolation and/or purification of certain components, etc. in some embodiments, the methods disclosed herein do not comprise a processed sample. Representative biological samples include, but are not limited to: blood, a component of blood, a portion of a tumor, plasma, serum, saliva, sputum, urine, cerebral spinal fluid, cells, a cellular extract, a tissue specimen, a tissue biopsy, or a stool specimen. In some embodiments a biological sample is whole blood and this whole blood is used to obtain measurements for a biomarker profile. In some embodiments a biological sample is tumor biopsy and this tumor biopsy is used to obtain measurements for a biomarker profile. In some embodiments a biological sample is some component of whole blood. For example, in some embodiments some portion of the mixture of proteins, nucleic acid, and/or other molecules (e.g., metabolites) within a cellular fraction or within a liquid (e.g., plasma or serum fraction) of the blood. In some embodiments, the biological sample is whole blood but the biomarker profile is resolved from biomarkers expressed or otherwise found in monocytes that are isolated from the whole blood. In some embodiments, the biological sample is whole blood but the biomarker profile is resolved from biomarkers expressed or otherwise found in red blood cells that are isolated from the whole blood. In some embodiments, the biological sample is whole blood but the biomarker profile is resolved from biomarkers expressed or otherwise found in platelets that are isolated from the whole blood. In some embodiments, the biological sample is whole blood but the biomarker profile is resolved from biomarkers expressed or otherwise found in neutrophils that are isolated from the whole blood. In some embodiments, the biological sample is whole blood but the biomarker profile is resolved from biomarkers expressed or otherwise found in eosinophils that are isolated from the whole blood. In some embodiments, the biological sample is whole blood but the biomarker profile is resolved from biomarkers expressed or otherwise found in basophils that are isolated from the whole blood. In some embodiments, the biological sample is whole blood but the biomarker profile is resolved from biomarkers expressed or otherwise found in lymphocytes that are isolated from the whole blood. In some embodiments, the biological sample is whole blood but the biomarker profile is resolved from biomarkers expressed or otherwise found in monocytes that are isolated from the whole blood. In some embodiments, the biological sample is whole blood but the biomarker profile is resolved from one, two, three, four, five, six, or seven cell types from the group of cells types consisting of red blood cells, platelets, neutrophils, eosinophils, basophils, lymphocytes, and monocytes. In some embodiments, a biological sample is a tumor that is surgically removed from the patient, grossly dissected, and snap frozen in liquid nitrogen within twenty minutes of surgical resection. [0048] A "score" is a numerical value that may be assigned or generated after normalization of the value based upon the presence, absence, or quantity of substrates or enzymes disclosed herein. In some embodiments, the score is normalized in respect to a control score.

[0049] The term "subject" is used throughout the specification to describe an animal from which a sample is taken. In some embodiment, the animal is a human. For diagnosis of those conditions which are specific for a specific subject, such as a human being, the term "patient" may be interchangeably used. In some instances in the description of the present invention, the term "patient" will refer to human patients suffering from a particular disease or disorder. In some embodiments, the subject may be a human suspected of having or being identified as at risk to develop a type of cancer more severe or invasive than initially diagnosed. In some embodiments, the subject may be diagnosed as having at resistance to one or a plurality of treatments to treat a disease or disorder afflicting the subject. In some embodiments, the subject is suspected of having or has been diagnosed with stage I, II, III or greater stage of cancer. In some embodiments, the subject may be a human suspected of having or being identified as at risk to a terminal condition or disorder. In some embodiments, the subject may be a mammal which functions as a source of the isolated sample of biopsy or bodily fluid. In some embodiments, the subject may be a non-human animal from which a sample of biopsy or bodily fluid is isolated or provided. The term "mammal" encompasses both humans and non-humans and includes but is not limited to humans, non-human primates, canines, felines, murines, bovines, equines, and porcines.

[0050] A "therapeutically effective amount" or "effective amount" of a composition

(e.g, any therapy or combination of therapies) is a predetermined amount calculated to achieve the desired effect, i.e., to improve and/or to decrease one or more symptoms of a disease or disorder. The activity contemplated by the present methods includes both medical therapeutic and/or prophylactic treatment, as appropriate. The specific dose of a compound administered according to this invention to obtain therapeutic and/or prophylactic effects will, of course, be determined by the particular circumstances surrounding the case, including, for example, the compound administered, the route of administration, and the condition being treated. The compounds are effective over a wide dosage range and, for example, dosages per day will normally fall within the range of from 0.001 to 10000 mg/kg, more usually in the range of from 0.01 to 1 mg/kg. However, it will be understood that the effective amount administered will be determined by the physician in the light of the relevant circumstances including the condition to be treated, the choice of compound to be administered, and the chosen route of administration, and therefore the above dosage ranges are not intended to limit the scope of the disclosure in any way. A therapeutically effective amount of compound of embodiments of this disclosure is typically an amount such that when it is administered in a physiologically tolerable excipient composition, it is sufficient to achieve an effective systemic concentration or local concentration in the tissue. In some embodiments the therapeutically effective amount is the amount of one or more active agents sufficient to treat or prevent cancers disclosed herein. In some embodiments, the therapeutically effective amount is the amount of one or more active agents sufficient to treat or prevent melanoma or a cancer derived from melanoma. In some embodiments, the therapeutically effective amount is the amount of one or more active agents sufficient to treat or prevent metastatic melanoma.

[0051] The terms "threshold value" as used herein refer to the quantitative value above which or below which a probability value is considered statistically significant as compared to a control set of data. For example, in the case of the disclosed method of determining the whether a nucleic acid pair corresponds to a likelihood of a subject or population of subjects to will be responsive to a therapy (such as therapy for subjects with melanoma), the threshold value is the quantitative value that is about 20%, 15%, 10%, 5%, 4%, 3%, 2%, or 1% below the greatest probability score assigned to a nucleic acid pair after the probability score is calculated by input of information associated with a disease or disorder into one or more of the statistical tests provided herein.

[0052] "Treatment" or "treating," as used herein can mean protecting of an animal from a disease or disorder through means of preventing, suppressing, repressing, or completely eliminating the disease or symptom of a disease or disorder. Preventing the disease involves administering a therapy (such as immune checkpoint blockade therapy, vaccine, antibody, biologic, gene therapy with or without viral vectors, small chemical compound, etc.) to a subject or population of subjects prior to onset of the disease or disorder. Suppressing the disease involves administering a therapy to a subject or population of subjects after induction of the disease but before its clinical appearance. Repressing the disease involves administering a therapy of to a subject or population of subjects after clinical appearance of the disease.

[0053] As used herein the term "web browser" means any software used by a user device to access the internet. In some embodiments, the web browser is selected from: Internet Explorer®, Firefox®, Safari®, Chrome®, SeaMonkey®, K-Meleon, Camino, OmniWeb®, iCab, Konqueror, Epiphany, Opera™, and WebKit®. [0054] Human or non-human variants of the enzymes above are contemplated by the methods, systems, and devices disclosed herein. Variants of these enzymes include sequences that are at least 70% homologous to the human sequences above. As used herein, the term "variants" is intended to mean substantially similar sequences. For nucleic acid molecules, a variant comprises a nucleic acid molecule having deletions (i.e., truncations) at the 5' and/or 3' end; deletion and/or addition of one or more nucleotides at one or more internal sites in the native polynucleotide; and/or substitution of one or more nucleotides at one or more sites in the native polynucleotide. As used herein, a "native" nucleic acid molecule or polypeptide comprises a naturally occurring nucleotide sequence or amino acid sequence, respectively. For nucleic acid molecules, conservative variants include those sequences that, because of the degeneracy of the genetic code, encode the amino acid sequence of one of the polypeptides of the disclosure. Variant nucleic acid molecules also include synthetically derived nucleic acid molecules, such as those generated, for example, by using site-directed mutagenesis but which still encode a protein of the disclosure. Generally, variants of a particular nucleic acid molecule of the disclosure will have at least about 70%, 75%, 80%, 85%, 87%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or more sequence identity to that particular polynucleotide as determined by sequence alignment programs and parameters as described elsewhere herein.

[0055] Variants of a particular nucleic acid molecule of the disclosure (i.e., the reference DNA sequence) can also be evaluated by comparison of the percent sequence identity between the polypeptide encoded by a variant nucleic acid molecule and the polypeptide encoded by the reference nucleic acid molecule. Percent sequence identity between any two polypeptides can be calculated using sequence alignment programs and parameters described elsewhere herein. Where any given pair of nucleic acid molecule of the disclosure is evaluated by comparison of the percent sequence identity shared by the two polypeptides that they encode, the percent sequence identity between the two encoded polypeptides is at least about 70%, 75%, 80%, 85%, 87%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or more sequence identity. In some embodiments, the term "variant" protein is intended to mean a protein derived from the native protein by deletion (so-called truncation) of one or more amino acids at the N-terminal and/or C-terminal end of the native protein; deletion and/or addition of one or more amino acids at one or more internal sites in the native protein; or substitution of one or more amino acids at one or more sites in the native protein. Variant proteins encompassed by the present disclosure are biologically active, that is they continue to possess the desired biological activity of the native protein as described herein. Such variants may result from, for example, genetic polymorphism or from human manipulation. Biologically active variants of a protein of the disclosure will have at least about 70%, 75%, 80%, 85%, 87%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or more sequence identity to the amino acid sequence for the native protein as determined by sequence alignment programs and parameters described elsewhere herein. A biologically active variant of a protein of the disclosure may differ from that protein by as few as 1-15 amino acid residues, as few as about 1 to about 10, such as 6-10, as few as about 20, 15, 10, 9, 8, 7, 6, 5, as few as 4, 3, 2, or even 1 amino acid residue. The proteins or polypeptides of the disclosure may be altered in various ways including amino acid substitutions, deletions, truncations, and insertions. Methods for such manipulations are generally known in the art. For example, amino acid sequence variants and fragments of the proteins can be prepared by mutations in the nucleic acid sequence that encode the amino acid sequence recombinantly.

[0056] The disclosure further relates to a computer program product encoded on a computer-readable storage medium that comprises instructions for performing any of the methods described herein. In some embodiments, the disclosure relates to any of the disclosed methods on a system or software that accesses the internet.

[0057] One application of such computers, computer program products, systems and methods is the identification of specific conditions for which a given chemical agent or one or more pharmaceutical drugs would provide effective therapeutic treatment. For example, the present invention provides systems and methods for identifying genetic expression profiles of specific cancers for which currently available chemical agents, pharmaceutical drugs, or other therapies of interest, such as immune checkpoint blockage therapy would provide either effective to treatment or ineffective due to resistance of treatment. The present disclosure also provides systems and methods for identifying genetic expression profiles of specific cancer susceptible to immune checkpoint blockage therapy for which currently available chemical agents, pharmaceutical drugs, would provide a therapeutically effective amount of a treatment or an adjuvant treatment. In some embodiments, the therapy is immune checkpoint blockage therapy.

[0058] The disclosure relates to a system or computer program product either integrated into a device or self-contained and, in either case, operably accessible to a database of expression information (mRNA expression or protein expression) of one or a plurality of nucleic acid sequences that are immune checkpoint proteins or variants thereof. In some embodiments, the computer program product of the disclosure comprises instructions execute any one or plurality of steps in any of the disclosed methods. In some embodiments, the computer program product, optionally integrated into a device, comprises instruction for executing the steps comprising: (a) analyzing information from a subject or population of subjects associated with a disease or disorder comprising a step selecting at least a first pair of nucleic acids comprising a first and second nucleic acid; (i) wherein expression of the first nucleic acid sequence and second nucleic acid sequence contributes to antigen- specific immune response against one or more cancer cells; and/or (ii) wherein mRNA or protein expression of both nucleic acid sequences contributes at least partially to a clinical outcome more positive than control subject or control population; and

(b) comparing a ratio of expression of at least the first pair of nucleic acid sequences with a survival rate associated with a disease or disorder in a control population of subjects; and

(c) assigning a probability score to the expression of at least the first pair of nucleic acid sequences based upon the survival rate of the subject or population of subjects diagnosed with or suspected as having cancer; and/or (d) selecting a therapy useful for treatment of the disease or disorder based upon the expression of at least the first pair of nucleic acid sequences.

In some embodiments, the disclosure provides systems and methods for defining and analyzing genetic profiles for at least one or two specific disease states (such as skin cancer); (2) identifying a therapy of interest (e.g., one or more chemical agents or one or more pharmaceutical drugs comprising immune checkpoint blockage therapy) known to be therapeutically effective in treating the specific disease state whose expression signature is defined by accessing and inputting information associated with the disease state or disorder from a database, (3) defining a discrimination set of genetic interactions that are

representative of changes in expression signatures or "response signature" for the genetic profile of the specific disease or disorder before, after administration of a therapy of interest induces a therapeutic effect; and (4) analyzing the screenable database to identify any other disease states that include a similar response signature for which the therapy of interest may be therapeutically effective in treating. The disclosure also relates to a computer program product accessible from a computer-usable or computer-readable medium providing program code for use by or in connection with a computer or any instruction execution system, including, in some embodiments, the aforementioned steps (1), (2), (3), and/or (4). In some embodiments, the computer program product is a component, at least in operative

communication with a processor of a device capable of measuring of expression levels of at least one or a plurality of nucleic acid sequences associated with immune checkpoint blockage therapy.

[0059] The disclosure relates to a method of quantifying the levels of antigen- specific immune activity in a subject with cancer or suspected of having cancer, a method of predicting clinical outcome or a method of predicting the responsiveness of a subject to immune checkpoint blockade therapy, wherein (a) the levels of antigen- specific immunity are the levels of antigen- specific immunity against the cancer or suspected cancer; (b) the predicting of clinical outcome; and/or (c) the predicting of responsiveness of a subject to a therapy comprises calculating a ratio of expression of nucleic acid sequences associated with or that are immune checkpoint proteins or variants expressed in a sample from a subject. In some embodiments, the ration of expression of the nucleic acids are a pair of a first and a second nucleic acid sequence, the first and second nucleic acid sequence comprising a sequence of at least about 70% homologous to any of SEQ ID NO: l, SEQ ID ΝΟ:2 ₅ SEQ ID NO:3, SEQ ID NO:4 ₅ SEQ ID NO:5 ₅ SEQ ID NO:6 ₅ SEQ ID NO:7, SEQ ID NO:8, SEQ ID NO:9, SEQ ID NO: 10, SEQ ID NO: 11, SEQ ID NO: 12, SEQ ID NO: 13, SEQ ID NO: 14, and/or SEQ ID NO: 15. In some embodiments, any of the mtethods disclosed herein comprise the step of acquiring, measuring or quantifying the expression of any one or plurality of immune checkpoint proteins or variants thereof prior to the step of calculating a ratio of expression of the nucleic acid sequences associated with immune checkpoint proteins or variants therof. In some embodiments, the value of expression for a first nucleic acid sequence or variant thereof is divided by the value of expression of a second nucleic acid sequence or variant therof. In some further embodiments, the step of calculating the ratio of expression comprises generating a pair of immune checkpoint proteins or variants thereof comprises normalizing the values of expression of the first and second nucleic acids or respective variants thereof prior to dividing their values. In some embodiments, the pair of genes expressing an immune check point protein or variants thereof comprises a first nucleic acid sequence (z) encoding an immune checkpoint protein selected from SEQ ID NO. 1 through SEQ ID NO: 15 or variants thereof and the second, different, nucleic acid sequence (j) encoding an immune checkpoint protein or variant thereof is also selected from SEQ ID NO. 1 through SEQ ID NO: 15 or variants thereof. Calulcation of any ratio of the mRNA or protein expression of the nucleic acids can be performed by the following formula I: where expi(x) and exp _j (x) denote the expression of checkpoint genes i and j in sample x. In any method disclosed herein, the average number Fi,j(x) can be calculated by adding the Fi,j(x) of each pair of nucleic acids and dividing that value by the number of pairs chosen to perform the analysis. In some embodiments, there are no fewer than 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 or more pairs of nucleic acids encoding encoding an immune checkpoint protein selected from SEQ ID NO. 1 through SEQ ID NO: 15 or variants thereof. In some embodiments, if the average Fi,i(x) level over the values of each ratio of nucleic acids is at or great than about 0.60, 0.65, 0.70, 0.75, 0.80, 0.85, 0.90, then that average value of Fi,j(x) can be used confirm that the subject has an antigen-specific immune response against the cancer that is characterized as high and thus predictive of the subject's good clinical outcome and/or responsiveness to immune checkpoint blockade therapy.

[0060] The pair of genes i and j in any of the methods disclosed herein can be any combination of two mRNA or protein expression values normalized against a control level of expression chosen from: SEQ ID NO: l through SEQ ID NO: 15 or a variant of any of the same comprising at least 70% sequence identity to any of SEQ ID NO: l through SEQ ID NO: 15. In some embodiments, any of the methods disclosed herein further comprise a step of determining the number of immune cells in a sample. In some embodiments, any of the methods disclosed herein further comprise the step of determining the number of immune cells in a sample comprises quantifying the number of CD8+ and CD4+ T cells. In any such embodiments, the numbers of CD8+ or CD4+ T cells can be used to correlate the level of antigen-specific immune activity against one or a plurality of cancer cells in the subject or can be used to correlate the likelihood that the subject responds to immune checkpoint blockade (ICB) therapy. The information may be then be used by a physician treating the subject to prescribe the subject ICB therapy or inform the subject that he or she has a good clinical outcome relative to the clinical outcome of a subject who is not responsive to an ICB therapy.

[0061] A good clinical outcome for a subject on ICB therapy, in some embodiments, may be about 30% relative improvement in 4-year overall survival rate (p < 0.001), as compared to subject noton the ICB therapy. In some embodiments, a good clincial outcome or a subject that responds to ICB thraepy demontsrates an improved median and 4-year OS of 16.9 months (95CI 15.6-19.3; vs. 7.7 months, 95CI 7.2-8.4) and 32.4% (95CI 29.5-35.3; vs. 21.0%, 95CI 19.6-22.2, all p < 0.001), respectively.

[0062] The disclosure also relates to a method of treating a subject in need thereof diagnosed or suspected of having melanoma, the subject having skin cancer and the method comprises: (a) calculating the ratio of expression of a first nucleic acid sequence over the expression of a second nucleic acid sequence in a sample from the subject; (b) assigning a probability score to at least the first pair of nucleic acid sequences based upon the ratio of expression of at least the first pair of nucleic acid sequences; and (c) identifying antigen- specific immune activity of the subject or the population of subjects based upon the ratio of expression of the first and the second nucleic acid sequences, wherein, if the ratio of expression of the first as compared to the second nucleic acid has an average score of above about 6, 7, 8, or 9 or greater, the subject is identified as having a high level of antigen- specific immune activity against the cancer; wherein the first nucleic acid sequence and the second nucleic acid sequence encode an immune checkpoint protein or variant thereof. In some embodiments, the method of treatment further comprises the step of determining the average probability score over at least about 5, 10, 11, 12, 13, 14, or 15 or more pairs of nucleic acid sequences and wherein the step of predicting the subject or population's responsiveness to the therapy is based upon the average probability score assigned over the at least 5, 10, 11, 12, 13, 14, or 15 pairs of nucleic acid sequences; and wherein, in some embodiments, the methods further comprise the step of administering to the subject in need thereof a pharmaceutical composition comprising a therapeutically effective amount of an immune checkpoint blockage therapy if the subject is identified as being responsive to ICB therapy or having a cancer susceptible to ICB therapy. In some embodiments, the methods further comprise a step of quantifying the levels of mRNA or protein expression of two or more immune checkpoint proteins or variants thereof in a sample from the subject, wherein the subject is diagnosed with having cancer or exhibiting one or more symptoms of cancer. Detection of mRNA or protein levels are known in the art and include in situ hybridization using antibodies capable of binding protein expressed by the nucleic acids disclosed herein, microarray analysis, PCR, RT-PCR, quantitative and semi-quantitiative PCR, fluorescence and/or absorbance readings of labels on the nucleic acids.

[0063] The disclosure relates to methods of identifying a level of antigen- specific immunity in a subject against at least one cancer cell in the subject comprising determining the level of RNA or protein expression of a first and second nucleic acid sequence in a sample from the subject, selecting the first and second nucleic acid sequences to form at least a first pair of nucleic acid sequences, calculating a score for relative expression as between the first and second nucleic acid sequence, an identifying the subject as having a high level of antigen-specific immunity against the at least one cancer cell if the score exceeds a threshold value related to a level of antigen-specific immunity in a control subject. In some embodiments, the methods of identifying a level of antigen-specific immunity in a subject against at least one cancer cell in the subject comprises selecting at least a first, second, third, fourth, fifth, and/or sixth nucleic acid sequence or more, and selecting at least a first, second, third or more pairs of the nucleic acid sequences. In some embodiments, the pair of nucleic acid sequences comprise two nucleic acid sequences associated with anti-CTLA-4 or anti- PD1 blockade therapy. In some embodiments, the methods disclosed herein comprises selecting, 1, 2, ,3 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 or more pairs of nucleic acid sequences, and a score corresponding to the relative expression ratio of the nucleic acid sequences in the is calculated. In some embodiments, the levels of antigen- specific immunity in the subject is based upon relative protein expression ratio of at least the first and second nucleic acid sequences. In some embodiments, the first and/or second nucleic acid sequences are based upon the expressible portion of nucleic acid sequences identified herein responsible for mRNA expression. In some embodiments, the first and/or second nucleic acid sequences are based upon the expressible portion of nucleic acid sequences identified herein responsible for protein expression. In some embodiments, the antigen- specific immunity is an antigen- specific immunity against a skin cell expession one or more tumor associated antigens. In some embodiments, the antigen-specific immunity in the subject is the level of immunity raised against one or a plurality of skin cancer cells. In some embodiments, the antigen- specific immunity in the subject is the level of immunity raised against a one or a plurality of tumor cells. In some embodiments, the antigen- specific immunity in the subject is the level of immunity raised against a one or a plurality of melanoma tumor cells. In some embodiments, the first or the second nucleic acid sequence encodes a protein or variant thereof associated with anti-immune checkpoint blockage therapy

[0064] The methods disclosed herein also relate to methods of determining a level of antigen-specific immunity in a subject diagnosed with cancer comprising calculating relative levels of expression of one or more pairs of nucleic acid sequences from a sample of the subject, calculating a probability score for each pair of nucleic acid sequences, and determining an average probability score over each pair of nucleic acids sequences, and correlating the average probability score to a threshold value such that if the score if at or higher than the threshold value, the subject is characterized as having a high level of antigen- specific immunity against the cancer of the subject and, if the average probability score is below the threshold value, the subject is characterized as not having a high antigen- specific immunity against the cancer of the subject. In some embodiments, the methods disclosed herein comprise determining relative ratio of expression of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 or more pairs of nucleic acid sequences that are immune checkpoint proteins or variants thereof and then calculating an average probability score over each pair of the nucleic acid sequences. Any pair of the disclosed nucleic acid sequences are possible in each embodiment, including any pairs disclosed in Supp. Table 8. In some embodiments, the first nucleic acid of the pair is the left-hand name of each pair disclosed in the columns of Supp. Table 8 and the second nucleic acid is the right-hand of name of each pair disclosed in Supp. Table 8.

[0065] The threshold values of any of the methods disclosed herein may be calculated by measuring the AUC of an ROC curve. ROC curve generation is known in the art. For instance, US Patent No. 7552035, which is incrorpoated by reference in its entirety, details a method of generating the curves with an acquired dataset. In some embodiments, and in any of the methods disclosed herein the step of calculating a probability score comprises generating an ROC curve and then calculating AUC for a given dataset. In some embodinets, the AUC may reflect the threshold value or the average AUC over several caluclations of expression of individual pairs (e.g. 5, 10, 11, 12, 13, 14 or 15 or more pairs) of nucleic acid sequence disclosed herein. Briefly, the present disclosure provides a method and system to determine the ability of a model to discriminate between normal operations or a fault situation may be evaluated using Receiver Operating Characteristic (ROC) curve analysis. In some embodiments, the step of calculating a probability score comprises performing an ROC analysis, calculating AUC and assigning a score based upon the AUC measurement in respect to the ratio of expression of each pair of nucleic acids analyzed and disclosed herein. The step of assigning a probability score can comprise, in such cases, comparing the average AUC values to a threshold value and characterizing the subject as being a responder to ICB therapy if the average AUC value is at or above the threshold value and characterizing the subject as not being a responder of ICB therapy if the average AUC value is lower than the threshold value. In some embodiments, the subject has metastatic cancer. When the results of a particular model are considered in two populations, one population without operating condition (e.g. responsive to ICB therapy), the other population operating with a fault (non-responsive to ICB therapy), there will rarely be a perfect separation between the two groups. For every possible cut-off point or criterion value that is selected to discriminate between the two populations, there will be some instances with the system operating normally and correctly classified as positive (TP=True Positive fraction), but some instances with the system operating normally that will be classified negative (FN=False Negative fraction). On the other hand, some cases operating with a fault will be correctly classified as negative (TN=True Negative fraction), but some cases operating with a fault will be classified as positive (FP=False Positive fraction).

[0066] The disclosure also relates to a method of determining the level of antigen- specific immunity of a subject to a cancer comprising correlating the one or plurality of probability scores to a particular clinical outcome. In some embodiments, being characterized as having a high antigen- specific immunity against the cancer is predictive of or correlated to the likelihood that the subject will respond to therapy, such as immune checkpoint blockade therapy. If the subject is characterized as being responsive to the therapy, the relative levels of each probability score can be used simultaneously to predict whether the prognosis of the subject is improved. Conversely if the subject is characterized as not having a high level of antigen-specific immunity, then the relative levels of each probability score can be used simultaneously to predict whether the prognosis of the subject is not improved.

[0067] The disclosure also relates to a method of detecting expression of immune checkpoint proteins or transcripts of the same in a sample of a subject with cancer or suspected as having cancer, the method comprising:

[0068] (a) contacting the sample with one or more probes specific to at least a pair of a first and a second nucleic acid sequence encoding a immune checkpoint protein or variant thereof;

(b) quantifying the expression of the immune checkpoint proteins by measuring the amount of probe non-covalently or covalently bound to at least the first and second nucleic acid sequences;

(c) calculating the ratio of expression of the first nucleic acid sequence over the expression of the second nucleic acid sequence in a sample from the subject;

(d) assigning a probability score to at least the first pair of nucleic acid sequences based upon the ratio of expression of at least the first pair of nucleic acid sequences; (e) identifying antigen- specific immune activity or ICB responsiveness of the subject or the population of subjects based upon the ratio of expression of the first and the second nucleic acid sequences. In some embodiments, the method comprises at least 5, 10, 11, 12, 13, 14, or 15 pairs of nucleic acid sequence combinations from the set of nucleic acids chosen from: SEQ ID NO: l through SEQ ID NO: 15. In some embodiments, the pairs of nucleic acids are those pairs identified in Supp. Table 8, whereby the first nucleic acid appears first in the pair and the second nucleic acid sequence appears second in the pair disclosed in Supp. Table 8. In some embodiments, the step of assigning a probability score comprises subjecting the one or more ratio of expression calculations for each nucleic acid pair to ROC analysis and then calculating the AUC of the ROC curve for each nucleic acid pair; and subsequently calculating the average AUC for the entire set of pair of nucleic acids. In some embodiments, the step of assigning further comprises assigning a 1 to any AUC calculation that is at about a value of 1 and assigning a 0 to any AUC calculation less than 1, and, if the average assigned value of the AUC is from about 1 or greater, the method comprises classifying the subject as having a high level of immune activity as compared to a control subject; and, if the average assigned value of the AUC is from about 0.99 or less, the method comprises classifying the subject as not having a high level of antigen- specific immune activity as compared to a control subject. In some embodiments, high levels of antigen-specific immune activity is correlated to the subject being a responder to ICB therapy and not having high levels of immune activity is correlated to the subject be a non-repsonder to ICB therapy. In some embodiments, the ratios of expression of at least one or several pairs of nucleic acid sequences is calculated by the Formula of page 45 of the application.

[0069] In any of the methods comprising a step of quantifying expression of the nucleic acid sequences encoding immune checkpoint protiens or variants thereof, the step may be accomplished by RNA sequencing, fluorescence, semi-quantitiative or quantitative PCR, absorbance measurements or the like. In some embodiments, the probe may be an antibody or an antibody probe that binds specifically to any of the proteins or variants thereof encoded by SEQ ID NO: l through 15.

[0070] In some embodiments, components and/or units of the devices described herein may be able to interact through one or more communication channels or mediums or links, for example, a shared access medium, a global communication network, the Internet, the World Wide Web, a wired network, a wireless network, a combination of one or more wired networks and/or one or more wireless networks, one or more communication networks, an a-synchronic or asynchronous wireless network, a synchronic wireless network, a managed wireless network, a non-managed wireless network, a burstable wireless network, a non-burstable wireless network, a scheduled wireless network, a non-scheduled wireless network, or the like.

[0071] Discussions herein utilizing terms such as, for example, "processing,"

"computing," "calculating," "determining," or the like, may refer to operation(s) and/or process(es) of a computer, a computing platform, a computing system, or other electronic computing device, that manipulate and/or transform data represented as physical (e.g., electronic) quantities within the computer's registers and/or memories into other data similarly represented as physical quantities within the computer's registers and/or memories or other information storage medium that may store instructions to perform operations and/or processes.

[0072] Some embodiments may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment including both hardware and software elements. Some embodiments may be implemented in software, which includes but is not limited to firmware, resident software, microcode, or the like.

[0073] Furthermore, some embodiments may take the form of a computer program product accessible from a computer-usable or computer-readable medium providing program code for use by or in connection with a computer or any instruction execution system. For example, a computer-usable or computer-readable medium may be or may include any apparatus that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.

[0074] In some embodiments, the medium may be or may include an electronic, magnetic, optical, electromagnetic, InfraRed (IR), or semiconductor system (or apparatus or device) or a propagation medium. Some demonstrative examples of a computer-readable medium may include a semiconductor or solid state memory, magnetic tape, a removable computer diskette, a Random Access Memory (RAM), a Read-Only Memory (ROM), a rigid magnetic disk, an optical disk, or the like. Some demonstrative examples of optical disks include Compact Disk-Read-Only Memory (CD-ROM), Compact Disk-Read/Write (CD- R/W), DVD, or the like.

[0075] In some embodiments, a data processing system suitable for storing and/or executing program code may include at least one processor coupled directly or indirectly to memory elements, for example, through a system bus. The memory elements may include, for example, local memory employed during actual execution of the program code, bulk storage, and cache memories which may provide temporary storage of at least some program code in order to reduce the number of times code must be retrieved from bulk storage during execution.

[0076] In some embodiments, input/output or I/O devices (including but not limited to keyboards, displays, pointing devices, etc.) may be coupled to the system either directly or through intervening I/O controllers. In some embodiments, network adapters may be coupled to the system to enable the data processing system to become coupled to other data processing systems or remote printers or storage devices, for example, through intervening private or public networks. In some embodiments, modems, cable modems and Ethernet cards are demonstrative examples of types of network adapters. Other suitable components may be used.

[0077] Some embodiments may be implemented by software, by hardware, or by any combination of software and/or hardware as may be suitable for specific applications or in accordance with specific design requirements. Some embodiments may include units and/or sub-units, which may be separate of each other or combined together, in whole or in part, and may be implemented using specific, multi-purpose or general processors or controllers. Some embodiments may include buffers, registers, stacks, storage units and/or memory units, for temporary or long-term storage of data or in order to facilitate the operation of particular implementations .

[0078] Some embodiments may be implemented, for example, using a machine- readable medium or article which may store an instruction or a set of instructions that, if executed by a machine, cause the machine to perform a method and/or operations described herein. Such machine may include, for example, any suitable processing platform, computing platform, computing device, processing device, electronic device, electronic system, computing system, processing system, computer, processor, or the like, and may be implemented using any suitable combination of hardware and/or software. The machine- readable medium or article may include, for example, any suitable type of memory unit, memory device, memory article, memory medium, storage device, storage article, storage medium and/or storage unit; for example, memory, removable or non-removable media, erasable or non-erasable media, writeable or re-writeable media, digital or analog media, hard disk drive, floppy disk, Compact Disk Read Only Memory (CD-ROM), Compact Disk Recordable (CD-R), Compact Disk Re- Writeable (CD-RW), optical disk, magnetic media, various types of Digital Versatile Disks (DVDs), a tape, a cassette, or the like. The instructions may include any suitable type of code, for example, source code, compiled code, interpreted code, executable code, static code, dynamic code, or the like, and may be implemented using any suitable high-level, low-level, object-oriented, visual, compiled and/or interpreted programming language, e.g., C, C++, Java, BASIC, Pascal, Fortran, Cobol, assembly language, machine code, or the like.

[0079] Functions, operations, components and/or features described herein with reference to one or more embodiments, may be combined with, or may be utilized in combination with, one or more other functions, operations, components and/or features described herein with reference to one or more other embodiments, or vice versa.

[0080] In one embodiment, the methods of this invention can be implemented by use of kits. Such kits contain software and/or software systems, such as those described herein. In some embodiments, the kits may comprise microarrays comprising a solid phase, e.g., a surface, to which probes are hybridized or bound at a known location of the solid phase. Preferably, these probes consist of nucleic acids of known, different sequence, with each nucleic acid being capable of hybridizing to an RNA species or to a cDNA species derived therefrom. In a particular embodiment, the probes contained in the kits of this invention are nucleic acids capable of hybridizing specifically to nucleic acid sequences derived from RNA species in cells collected from subject of interest. In some embodiments, any of the disclosed methods comprise a step of obtaining or providing information associated with a disease or disorder. In some embodiments, the step of obtaining or providing comprises isolating a sample from a subject or population of subjects and, optionally performing a genetic screen to obtain expression data or nucleic acid sequence activity data which can then be analyzed with other disclosed steps as compared to a control subject or control population of subjects.

[0081] In some embodiments, data or information associated with a subject or population of subjects may be obtained by an individual patient and scored across any or all of the steps disclosed herein by comparing the analysis to information associated with a disease or disorder from a control subject or control population of subjects. In some embodiments, the disease is cancer. In some embodiments, the data or information associated with a disease is taken from any of the data provided in https://gdc-portal.nci.nih.gov, an NIH database of clinical data, which is hereby incorporated by reference in its entirety. Any of the data from the website may be analyzed across one or a plurality of conditions including cancer types disclosed on within the NIH database.

[0082] In some embodiments, a kit of the invention also contains one or more databases described above, encoded on computer readable medium, and/or an access authorization to use the databases described above from a remote networked computer. [0083] In another embodiment, a kit of the invention further contains software capable of being loaded into the memory of a computer system such as the one described above. The software contained in the kit of this invention, is essentially identical to the software described above.

[0084] Alternative kits for implementing the analytic methods of this invention will be apparent to one of skill in the art and are intended to be comprehended within the accompanying claims.

[0085] Although the disclosure has been described with reference to exemplary embodiments, it is not limited thereto. Those skilled in the art will appreciate that numerous changes and modifications may be made to the preferred embodiments of the disclosure and that such changes and modifications may be made without departing from the true spirit of the disclosure. It is therefore intended that the appended claims be construed to cover all such equivalent variations as fall within the true spirit and scope of the disclosure.

[0086] Any and all journal articles, patent applications, genelD references, websites or other GenBank or Accession Numbers disclosed herein are hereby incorporated by reference in their entireties.

EXAMPLES

Example 1; Methods

Statistical analyses

[0087] Boxplots and comparisons. For all boxplots, center lines indicate medians, box edges represent the interquartile range, whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted individually using the '+' symbol. Points are defined as outliers if they are greater than q^ + w x ( 3 - q\) or less than qi - w x ( 3 - qi), where w is the maximum whisker length, and q _\ and ^3 are the 25th and 75th percentiles of the sample data, respectively. All differential expression and distribution comparisons P- values are obtained via one-sided Rank-sum test.

[0088] Survival analyses. All Kaplan Meier analyses are performed by comparing the survival of patients with high scores (> median) to those with low scores (< median) using a two-sided log-rank test. The patients with median score (= median) are grouped with the smaller-size group among the two groups mentioned above.

[0089] Bar plots. For bar plots, center is defined by the mean of the distribution and error bars represents the SD of the distribution. [0090] Correlation coefficients. All correlations coefficients and P-values are obtained via Spearman rank correlation test.

Collection of immune checkpoint molecules

[0091] To build a predictor based on pair-wise relations between checkpoint genes' expression, we formed a list of 45 immune checkpoint genes with known co-stimulatory or co inhibitory effects, collected from literature reports 21—24. From these, we focus on 28 genes that were measured in all RN A- sequencing datasets analyzed in this paper (See Supp. Table 1 below)

Supp. Table \.

Cover

Activator( ed

A) Datas

Checkpoin Inhibitor(I VanAll Hug TCG Che Pr Ria ets

t gene ) en o A n at z count Reference

ICOS A 0 1 1 1 1 1 5 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

MC3786574/

ICOSL A 0 0 0 0 0 0 0 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

MC3786574/

IDOl I 1 1 1 1 1 1 6 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

MC3786574/

http ://w ww .cell . com immunity/pdf/S 1074-

IL2RB A 1 1 1 1 1 1 6 7613(16)30146-7.pdf

KIR3DL http ://w ww .cell .com immunity/pdf/S 1074- 1 I 0 0 1 0 1 1 3 7613(16)30146-7.pdf

LAG3 I 1 1 1 1 1 1 6 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

MC3786574/

NAIL A 0 0 0 0 0 0 0 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

MC3786574/

LAIR1 I 1 0 1 0 0 1 4 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

MC3786574/

PD-1 I 1 1 1 1 1 1 6 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

MC3786574/

PD- I 1 1 1 1 1 1 6 https://www.ncbi.nlm.nih.gOv/pmc/articles/P 1LG2 MC3786574/

http ://w ww .cell .com immunity/pdf/S 1074-

PVR I 1 1 1 0 1 1 5 7613(16)30146-7.pdf

http ://w ww .cell .com immunity/pdf/S 1074-

PVRL2 I 1 1 1 0 0 1 4 7613(16)30146-7.pdf

https://www.ncbi.nlm.nih.gOv/pmc/articles/P

SLAM A 0 0 0 0 0 0 0 MC3786574/

TIGIT I 1 1 1 0 1 1 5 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

MC3786574/

TIM2 A 0 0 0 0 0 0 0 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

MC3786574/

HVEM A 1 1 1 1 1 6 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

¹ MC3786574/

TNFRS A 1 1 1 1 1 6 https://www.ncbi.nlm.nih.gOv/pmc/articles/P F18 ¹ MC3786574/

TNFRS A 1 1 1 1 1 6 https://www.ncbi.nlm.nih.gOv/pmc/articles/P F4 ¹ MC3786574/

TNFRS A 1 1 1 1 1 6 https://www.ncbi.nlm.nih.gOv/pmc/articles/P F9 ¹ MC3786574/

TNFSF1 A 0 1 1 1 1 5 https://www.ncbi.nlm.nih.gOv/pmc/articles/P 4 ¹ MC3786574/

TNFSF1 A 0 0 1 1 1 4 https://www.ncbi.nlm.nih.gOv/pmc/articles/P 8 ¹ MC3786574/

OX40L A 1 1 1 1 1 6 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

¹ MC3786574/

CD137L A 1 1 1 0 1 5 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

¹ MC3786574/

VTCN1 I 0 0 1 0 0 1 2 https://www.ncbi.nlm.nih.gOv/pmc/articles/P

MC3786574/ Feature selection and IMPRES construction on the NB data

[0092] For feature selection, we use the quantile-normalized expression of the 28 immune checkpoint genes selected above in the 108 NB tumor samples studied, using the following expression function of pairs of checkpoint genes as features:

^1,1 lo, otherwise, Where exp _t (x) and exp _j (x) denote the expression of genes i and j in sample x.

[0093] We focus on pairs where at least one of the genes is among the six genes that are directly associated with anti-CTLA-4 and anti-PDl blockade therapy, including CTLA-4, CD28, CD80/CD86, PD-1 and PD-L1 ²⁵ , which together form 294 potential gene pairs. To select features that best separate positive from negative samples in the NB data, we performed a hill climbing aggregative feature selection involving 500 iterations of a five-fold cross validation procedure, where the features that highly scored consistently across folds were selected for IMPRES. A detailed description of the feature selection steps is available below.

Immune pathway enrichment analysis

[0094] To identify CDPs (consistently differentially expressed immune pathways in melanoma ICB responders), we first identified the genes that are up and down regulated in ICB responders vs non-responders for each of the datasets ¹ ' ³ ' ⁴ ' ⁶ (using one sided Rank-sum P- value<0.05). Then, we performed a GO pathway ²⁶ enrichment analysis for immune related pathways (data not shown) via a hyper-geometric test, to identify (1) pathways that are consistently up or down regulated (hyper-geometric P-value<0.05) in responders for all anti- PD-1 melanoma datasets, and (2) pathways that are consistently up or down regulated in responders for all anti-CTLA-4 melanoma datasets (Figure 1C, Supp. Table 3).

[0095] To correlate CDPs with the IMPRES features, we then evaluated the

Spearman rank correlation coefficients (p) and corresponding P-values between the median pathway expression level of each CDP (using the median expression of all genes in a pathway) and each of the IMPRES expression ratios. This is done across all samples in each of the following datasets: (1) the anti-PD-1 treated melanoma datasets (2) the anti-CTLA-4 treated melanoma datasets (3) the non ICB -treated melanoma datasets and (4) the neuroblastoma dataset. Computing IMPRES features' expression ratio

[0096] To evaluate the predictive performance and functional associations of individual IMPRES features in a more refined manner we used the expression ratio instead of the binary indicators in each sample (i.e. for each feature A>B we used A/B instead). The resulting AUCs obtained with each ratio feature for each ICB response data are presented in Figure 3D and Supp. Table 7B (abridged from original) appears immediately below wherein the rows of the second table labeled "Score" correspond to the rows of the first table labeled "Feature or Pair of Nucleic Acid".

0.51 71 97 0.543835 0.501 537791 0.555226481

0.330648 0.4421 1 6 0.361 1 86101 0.256484708

0.581 7 0.5935 0.549 0.661 2

CIBERSORT analysis

[0097] Using CIBERSORT 27 we infer the relative abundances of 22 immune cells in

NB and melanoma samples (analyzing each dataset for ICB treated and non-treated melanoma). Having estimated these cell abundances for each sample, we perform the following analyses:

1. We perform a differential abundance analysis via a one-sided Rank-sum test for each immune cell type between NB samples with or without spontaneous regression, between melanoma samples who respond or do not respond to ICB treatment, and between high immune response versus other subtypes in non ICB-treated melanoma samples (Supp. Table 4).

2. To study the relation between IMPRES and major T cell types mediating the immune response, we correlate the CIBERSORT inferred relative abundances of CD8+ and CD4+ T-cells to IMPRES scores for RNA-seq melanoma ICB response datasets via Spearman rank-correlation (Figure 7).

3. To survey the key associations between IMPRES features and immune subtypes, we correlate each CIBERSORT inferred immune cell type to each IMPRES feature (considering expression ratios instead of binary relations), for RNA-seq melanoma ICB response datasets via Spearman rank-correlation (data not shown).

Applying IMPRES to predict ICB response of melanoma patients

[0098] To apply IMPRES, we calculate for each sample x, the F _{i ;} (x) over the 15

IMPRES checkpoint pairs (features). This leads to a binary vector of length 15 for each sample. The total number of Ts in this vector denotes the sample's IMPRES score (ranging between 0 and 15). High scores predict good response. By varying the classification threshold over the different possible IMPRES score values we generate the ROC curves and the resulting AUCs presented in the main text for each melanoma dataset.

RNA-seq

[0099] RN A- sequencing of 31 anti-PD-1 pre- and on-treatment tumor specimens, and 10 anti-CTLA-4 pre- and on- treatment metastatic tumor specimens (for which the response is known) derived from patients with metastatic melanoma (up to 90 days from treatment start) was conducted as previously described in Jenkins et al 28 (Supp. Table 9). These patients were enrolled in clinical trials at Massachusetts General Hospital. Clinical trial registration numbers at ClinicalTrials.gov are NCT01714739; NCT02083484; NCT01543698; NCT01072175; NCT00949702; NCT01783938; NCT01006980.

Clinical response classification

[00100] Table 1 below summarizes the response annotations and criteria used for establishing them in the original study. The response classification of each patient in each of the publicly available studies and the MGH dataset (with MGH patients clinical information) is not shown but in N. Alexander, et al. Nature Medicine, 2018, incorporated by reference in its entirety.

Table 1. Response annotations for each melanoma dataset

Applying IMPRES to predict melanoma patients' response to ICB treatments on the combined, aggregate collection of all melanoma samples

[00101] To apply IMPRES and evaluate its predictive performance on the combination of melanoma samples in all eleven datasets studied, we normalize the IMPRES scores for datasets in which not all relevant checkpoint genes were measured (Supp. Table 1). This normalization is done by linearly scaling the IMPRES score to compensate for the number of missing pairs whose expression is not available in the dataset. For example, if 13 out of the 15 IMPRES features are measured in a dataset, then the IMPRES score of these samples is multiplied by 15 and divided by 13 to linearly scale it back to the original scale of 0-15. We then calculate the AUC in a standard manner using these normalized IMPRES scores over all samples. Training a predictor using melanoma datasets

[00102] We use a similar training procedure as described above for IMPRES (using binary relations (A>B) between the 28 checkpoint genes as before), but this time training on melanoma datasets: (1) First we train on the combined data from Riaz et al ⁶ and Hugo et al ⁹ (both anti-PD-1 datasets) and (2) we then train on the combined datasets from Hugo et al. and

Van Allen et al. 1 ' 3 (anti-PD-1 and anti-CTLA-4 datasets, respectively). For both, we use hill climbing feature selection and perform similar procedure as described above in 'Feature selection and IMPRES construction on the NB data', of 500 rounds of five-fold training and testing. The final feature set is also selected in the same manner (using similar definition of score(f) and selecting features with binomial p-value < 0.05).

IMPRES analysis of different melanoma subtypes

[00103] We evaluate whether IMPRES significantly differs between different melanoma subtypes by comparing IMPRES scores of each subtype against that of all other subtypes, when using pre anti-PD-1, on anti-PD-1 and all samples from Riaz et al (Figure 10)

6

Comparing IMPRES predictive performance to that obtained by predictors based on other published signatures

[00104] We compare the performance of IMPRES to those obtained using other published transcriptomic signatures 3 ' 15 ' 31 as well as PDL-1 expression. We additionally evaluate the performance of a predictor based on immune cell abundances estimated via

CIBERSORT 27. The predictors' performance is evaluated using the nine publicly available melanoma datasets analyzed to evaluate IMPRES (see Example 2). The Cytolytic activity ¹⁵ and PDL1 expression based predictors are applied in a straightforward manner, analogous to that of IMPRES as they do not require additional training. However, making predictions using gene signatures reported in specific studies in the literature (see Example 2) requires training on every specific dataset tested (using cross validation), aiming to identify their maximal performance levels. Hence, we build predictors of ICB response using Support Vector Machines (SVMs) on each of the pertaining melanoma datasets. Each such SVM predictor is built using the genes in the specific signatures on which it is based as its feature set. This is performed with linear kernels using 100 repetitions of a five-fold cross validation process, where in each fold the training set and test set are randomly selected. The AUC presented for each predictor is the mean AUC overall repetitions (Supp. Table 6, upper panel).

Supplementary Table 6. Comparison of IMPRES performance to that of other classifiers (AUC obtained by each classifier for each dataset)

TCGA 0.5513 0.665 0.6325 0.925 0.5333 0.4749 0.55 0.8167

SKCM

Chen et 1 0.835 1 0.7975 1 0.5007 1 0.96 al. - On

aPDl

Chen et 0.4725 0.5038 0.225 0.4837 0.5873 0.7147 0.5818 0.7698 al. - pre

aPDl

Chen et

al. - pre

aCTLA4 0.385 0.48 0.445 0.4875 0.6364 0.721 0.5397 0.8

Felip et 0.5061 0.4832 0.5555 0.3925 0.6753 0.5581 0.6169 0.7305 al.

Riaz et 0.4775 0.52 0.5757 0.5606 0.6481 0.312 0.6463 0.8 al. pre

aPD-1

Riaz et 0.4875 0.19 0.1774 0.3587 0.7623 0.4278 0.7522 0.916 al. On

aPD-1

0.53053567 0.44138 0.72195513 0.69215 0.83453

All aPDl 6 0.444058378 7027 0.54088 5 0.475645946 4595 8919

All 0.47348028 0.44011 0.67556 0.66360422 0.58240 0.82033 aCTLA4 2 0.448447887 1268 338 5 0.634235211 2817 0986

All pre- 0.51028 0.59898 0.63819508 0.64083 0.78886 treatment 0.51075082 0.522383607 1967 9344 2 0.494377869 6066 6393

All on- treatment 0.56884920 0.30797 0.42835 0.88415714 0.79153 0.92298

PD1 6 0.292380952 1429 0794 3 0.439371429 3333 4127

0.44103 0.57823 0.70577187 0.66171 0.83059

All Data 0.514711719 0.445275781 3203 3594 5 0.519629688 5625 8438

[00105] To compare IMPRES performance to that of other predictors over different treatment groups in a systematic manner, we aggregate the samples into four treatment groups: pre anti-CTLA-4, pre anti-PD-1, on anti-PD-1 and all samples. To calculate an empirical P-value, we then perform 1000 repetition of: (1) randomly sampling 80% of the samples in a stratified manner (maintaining the proportion of responders vs. non-responders) from each treatment group and (2) evaluating the AUC resulting from each predictor on the randomly selected samples from each treatment group. The resulting empirical P-value (reported in Figure 3B) denotes the percentage of iterations in which the AUC obtained via each predictor is superior to that obtained using IMPRES for each treatment group.

Feature reduction analysis

[00106] As IMPRES features were selected for NB, it is possible that some are less predictive for melanoma response. To investigate which features may be removed, we performed a feature removal procedure using 5-fold cross validation on the combined data from all 5 RNAseq datasets ^ ³ ' ⁴ ' ⁶ (as these cover all IMPRES features). In each fold we performed 500 rounds of greedy (hill climbing) feature removal on the training set, each round ended when the performance (AUC) has been reduced by more than 5% from the full set performance on the training set. The set of remaining features was applied to the test set to obtain a test-AUC. We find a group of 11 features that captures most of IMPRES original performance levels (Figure 9, Supp. Table 8). All groups with less than 11 features have reduced the test-AUC by more than 5%.

Supplementary Table 8. Features used for classifiers trained on melanoma data, on both PCA clusters and via feature reduction procedure

IMPRES sensitivity analysis and random control predictors

[00107] To evaluate IMPRES sensitivity to missing features, we perform all possible removals of single, double, triple and quadruple features. For each removal, we examine the AUCs obtained with the remaining features for each of the 11 datasets. We find that while the results remain robust for most single and double feature removals, higher order removals are likely to significantly reduce the performance but not in all cases (e.g., as the reduced features set, Figure 9, Supp. Table 10).

Supplementary Table 10 Response annotations for each melanoma dataset

"response 'Partial 'Partial decreased tumor

Response' Response/ > 6 months

classified 'nonrespon 'Progressiv 'Clinical 'NR' = tumor 'PD' 'PD', 'SD' 'PD' as "non- se' , 'long- e Disease' Progressive growth on serial

response" survival' Disease', CT scans or a

'Stable clinical benefit

Disease' lasting 6 months

or less

Protocol irPvECIST ² RECIST ³⁰ RECIST ³⁰ Nan RECIST ³ RECIST ³ RECIST ³

[00108] To evaluate the power of predictors constructed via randomly selected relations, we generate 1000 predictors, each based on randomly selected 15 immune gene relations (drawing from the same pool of features as those considered for the construction of IMPRES, Figure 11).

Principle Component Analysis (PCA) using IMPRES features

[00109] We perform PCA of patient profiles for each melanoma ICB study and across all studies combined using IMPRES selected features (i.e., each sample is represented as a 15-dimensional binary vector comprised of the 15 IMPRES logical relations). The PCA results are presented in Figure 12.

Comparing IMPRES scores to mutational counts across TCGA tumors

[00110] Examining pan-cancer TCGA data, we find that cancer types with high

IMPRES scores tend to have a higher mutational burden, a well-established marker of response to immunotherapy (Spearman Rho = 0.79, Figure 13). To perform this analysis, the complete raw data of all TCGA cancer types (n=23) which include at least 100 patients were downloaded from cBioPortal 32. We considered samples containing somatic point mutations and gene expression data, relative to matched-normal samples.

Code availability

[00111] Codes are implemented in MATLAB and are publicly available in GitHub: https://github.com/noamaus/IMPRES-codes

[00112] The CIBERSORT software was applied on matrices of gene expression data using the given LM22 immune cell signatures.

Data availability statement

[00113] All patients data analyzed from published papers are referenced to and publicly available accordingly. The transcriptomic data of the MGH patients analyzed are available from GEO (GSE115821). MGH patients' clinical information is available in Supp. Table 9.

Normalization and preparation of the neuroblastoma data

[00114] The neuroblastoma (NB) data ⁹ includes a full set of gene expression profiles of 498 NB samples with the following clinical features: (1) High risk: clinically considered as high-risk neuroblastoma (yes=l, no= 0) (2) Progression: Occurrence of a tumor progression event (yes=l; no=0), (3) Class label: Maximally divergent disease courses - unfavorable (= 1): patient died despite intensive chemotherapy, favorable (=0): patient survived without chemotherapy for at least 1000 days post diagnosis; not applicable (N/A). Class label is available only for a subset of the NB patients, and we hence use the 'High risk' and 'Progression' annotations for the analysis (data not shown) .

[00115] We first perform principle component analysis (PCA) with all 498 samples.

We note that while PCI (explaining 7.2% of the variance) shows differences in clinical annotations of the samples (between those considered 'high risk' vs 'not high risk' and those considered 'favorable disease course' vs. 'not favorable disease course', Figure 4A). PC2 and PC3 (explaining 3.8% and 2.3% of the variance, respectively) reveal two different clusters that are not associated with any clinical features.

[00116] To perform feature selection, we focus on analyzing patients younger than 18 months, as spontaneous regression cases have almost exclusively been observed in this age group while patients older than 18 months have tumors that are mostly unresectable or metastatic ^{10 13} . We separate the tumor samples of the patients younger than 18 months to two groups: (1) samples classified as 'not high risk' and with no cancer progression (spontaneous regression), and (2) samples classified as 'high risk' and with cancer progression (no spontaneous regression). This results with 226 samples, from which 209 are classified as "spontaneous regression."

[00117] We first perform feature selection using all 226 relevant samples, however, the top selected features (using the procedure described below) had only moderate performance when applied to all NB samples (i.e. AUC < 0.8 for all score f P-value thresholds evaluated (defined below), Figure 4B). As one of the clusters (defined by PC2+PC3<0) has only one sample in the "no spontaneous regression" class, it cannot be used for feature selection on its own. We thus turn to perform a similar feature selection procedure on the remaining cluster, defined by PC2+PC3>0. This cluster has in total 108 samples, 92 samples in the "spontaneous regression" class and 16 samples in the "no spontaneous regression" class. Analyzing this cluster, we find that the data becomes more separable and leads to higher prediction accuracy on the NB dataset than when considering all relevant samples, and that this remains robust across multiple score f P-value thresholds (employed for selecting the feature set, Supp. Figure IB). We hence report the results obtained analyzing the cluster defined by PC2+PC3>0 below and we report the results obtained when analyzing all NB samples in the section titled 'Feature selection using all NB data samples' .

Feature selection and IMPRES construction on the NB data

[00118] For feature selection, we use the quantile-normalized expression of the 28 immune checkpoint genes marked in Supp. Table 1 in the 108 NB tumor samples studied. To create a generalizable predictor that can then be transferred and applied to melanoma samples, we use the following pairwise function of these immune checkpoint genes as p. . x) = f ¹ ' ^e¾ i¾(*) ^{< ex} P ^j

^Ej {Q _f otherwise-,

features:

Where expi(x) and exp _j (x) denote the expression of checkpoint genes i and j in sample x.

[00119] Out of all possible pairwise permutations of the 28 checkpoint genes F/,1

(N=756), we focus on i, j pairs containing one of PD-1 and its ligand PDL-1, CTLA-4, its homologous receptor CD28 and their ligands CD80 and CD86, as those are the key checkpoint genes associated with either PD-l/PDL-1 or CTLA-4 blockade. This reduces the feature space to 294 potential pairs (See Supp. Table 2).

Supp. Table 2

Considered checkpoint pairs IMPRES (selected) checkpoint pairs

Gene 1 Gene 2 Gene 1 Gene 2

PD-1 CD200

PDL-1 CD200R1

CD28 CD200R1

CD80 CD200R1

CD86 CD200R1

CTLA4 CD200R1

PD-1 CD200R1

PDL-1 CD27

CD28 CD27

CD80 CD27

CD86 CD27

CTLA4 CD27

PD-1 CD27

CD276 PDL-1

CD28 PDL-1

CD40 PDL-1

CD80 PDL-1

CD86 PDL-1

CEACAM1 PDL-1

CTLA4 PDL-1

TIM-3 PDL-1

IDOl PDL-1

IL2RB PDL-1

LAG3 PDL-1

PD-1 PDL-1

PD-1LG2 PDL-1

PVR PDL-1

PVRL2 PDL-1

TIGIT PDL-1

HVEM PDL-1

TNFRSF18 PDL-1

TNFRSF4 PDL-1

TNFRSF9 PDL-1

OX40L PDL-1

CD137L PDL-1

CD28 CD276

CD80 CD276

CD86 CD276

CTLA4 CD276

PD-1 CD276

CD40 CD28

CD80 CD28

CD86 CD28

CEACAM1 CD28

CTLA4 CD28

TIM-3 CD28

IDOl CD28

IL2RB CD28

LAG3 CD28

PD-1 CD28

PD-1LG2 CD28

PVR CD28

PVRL2 CD28

TIGIT CD28

HVEM CD28

TNFRSF18 CD28 Considered checkpoint pairs IMPRES (selected) checkpoint pairs

Gene 1 Gene 2 Gene 1 Gene 2

TNFRSF4 CD28

TNFRSF9 CD28

OX40L CD28

CD137L CD28

CD80 CD40

CD86 CD40

CTLA4 CD40

PD-1 CD40

CD86 CD80

CEACAM1 CD80

CTLA4 CD80

TIM-3 CD80

IDOl CD80

IL2RB CD80

LAG3 CD80

PD-1 CD80

PD-1LG2 CD80

PVR CD80

PVRL2 CD80

TIGIT CD80

HVEM CD80

TNFRSF18 CD80

TNFRSF4 CD80

TNFRSF9 CD80

OX40L CD80

CD137L CD80

CEACAM1 CD86

CTLA4 CD86

TIM-3 CD86

IDOl CD86

IL2RB CD86

LAG3 CD86

PD-1 CD86

PD-1LG2 CD86

PVR CD86

PVRL2 CD86

TIGIT CD86

HVEM CD86

TNFRSF18 CD86

TNFRSF4 CD86

TNFRSF9 CD86

OX40L CD86

CD137L CD86

CTLA4 CEACAM1

PD-1 CEACAM1

TIM-3 CTLA4

IDOl CTLA4

IL2RB CTLA4

LAG3 CTLA4

PD-1 CTLA4

PD-1LG2 CTLA4

PVR CTLA4

PVRL2 CTLA4

TIGIT CTLA4

HVEM CTLA4

TNFRSF18 CTLA4 Considered checkpoint pairs IMPRES (selected) checkpoint pairs

Gene 1 Gene 2 Gene 1 Gene 2

TNFRSF4 CTLA4

TNFRSF9 CTLA4

OX40L CTLA4

CD137L CTLA4

PD-1 ΤΓΜ-3

PD-1 IDOl

PD-1 IL2RB

PD-1 LAG3

PD-1LG2 PD-1

PVR PD-1

PVRL2 PD-1

TIGIT PD-1

HVEM PD-1

TNFRSF18 PD-1

TNFRSF4 PD-1

TNFRSF9 PD-1

OX40L PD-1

CD137L PD-1

BTLA PDL-1

BTLA CD28

BTLA CD80

BTLA CD86

BTLA CTLA4

BTLA PD-1

VISTA PDL-1

VISTA CD28

VISTA CD80

VISTA CD86

VISTA CTLA4

VISTA PD-1

CD200 PDL-1

CD200 CD28

CD200 CD80

CD200 CD86

CD200 CTLA4

CD200 PD-1

CD200R1 PDL-1

CD200R1 CD28

CD200R1 CD80

CD200R1 CD86

CD200R1 CTLA4

CD200R1 PD-1

CD27 PDL-1

CD27 CD28

CD27 CD80

CD27 CD86

CD27 CTLA4

CD27 PD-1

PDL-1 CD276

PDL-1 CD28

PDL-1 CD40

PDL-1 CD80

PDL-1 CD86

PDL-1 CEACAM1

PDL-1 CTLA4

PDL-1 TIM-3 Considered checkpoint pairs IMPRES (selected) checkpoint pairs

Gene 1 Gene 2 Gene 1 Gene 2

PDL-1 IDOl

PDL-1 IL2RB

PDL-1 LAG3

PDL-1 PD-1

PDL-1 PD-1LG2

PDL-1 PVR

PDL-1 PVRL2

PDL-1 TIGIT

PDL-1 HVEM

PDL-1 TNFRSF18

PDL-1 TNFRSF4

PDL-1 TNFRSF9

PDL-1 OX40L

PDL-1 CD137L

CD276 CD28

CD276 CD80

CD276 CD86

CD276 CTLA4

CD276 PD-1

CD28 CD40

CD28 CD80

CD28 CD86

CD28 CEACAM1

CD28 CTLA4

CD28 TIM-3

CD28 IDOl

CD28 IL2RB

CD28 LAG3

CD28 PD-1

CD28 PD-1LG2

CD28 PVR

CD28 PVRL2

CD28 TIGIT

CD28 HVEM

CD28 TNFRSF18

CD28 TNFRSF4

CD28 TNFRSF9

CD28 OX40L

CD28 CD137L

CD40 CD80

CD40 CD86

CD40 CTLA4

CD40 PD-1

CD80 CD86

CD80 CEACAM1

CD80 CTLA4

CD80 TIM-3

CD80 IDOl

CD80 IL2RB

CD80 LAG3

CD80 PD-1

CD80 PD-1LG2

CD80 PVR

CD80 PVRL2

CD80 TIGIT

CD80 HVEM Considered checkpoint pairs IMPRES (selected) checkpoint pairs

Gene 1 Gene 2 Gene 1 Gene 2

CD80 TNFRSF18

CD80 TNFRSF4

CD80 TNFRSF9

CD80 OX40L

CD80 CD137L

CD86 CEACAM1

CD86 CTLA4

CD86 TIM-3

CD86 IDOl

CD86 IL2RB

CD86 LAG3

CD86 PD-1

CD86 PD-1LG2

CD86 PVR

CD86 PVRL2

CD86 TIGIT

CD86 HVEM

CD86 TNFRSF18

CD86 TNFRSF4

CD86 TNFRSF9

CD86 OX40L

CD86 CD137L

CEACAM1 CTLA4

CEACAM1 PD-1

CTLA4 TIM-3

CTLA4 IDOl

CTLA4 IL2RB

CTLA4 LAG3

CTLA4 PD-1

CTLA4 PD-1LG2

CTLA4 PVR

CTLA4 PVRL2

CTLA4 TIGIT

CTLA4 HVEM

CTLA4 TNFRSF18

CTLA4 TNFRSF4

CTLA4 TNFRSF9

CTLA4 OX40L

CTLA4 CD137L

TIM-3 PD-1

IDOl PD-1

IL2RB PD-1

LAG3 PD-1

PD-1 PD-1LG2

PD-1 PVR

PD-1 PVRL2

PD-1 TIGIT

PD-1 HVEM

PD-1 TNFRSF18

PD-1 TNFRSF4

PD-1 TNFRSF9

PD-1 OX40L

PD-1 CD137L [00120] To choose the features that best separate spontaneous regression from no spontaneous regression samples in the NB data, we use a hill climbing aggregative feature selection procedure. This procedure starts from an empty set and incrementally adds the best discriminating feature at each step in a greedy manner, until no further improvement in the prediction accuracy is obtained. More formally, we repeat the following procedure with 500 iterations:

Randomly select 13 samples from each class (spontaneous regression or no spontaneous regression) as training set, and 3 samples in each class as test set.

Initialize the current group of features to the empty set and set the current training- AUC to 0.

While the current training-AUC<l:

Calculate the training-AUC resulting from adding each feature to the current group of features one at a time

Select the feature that maximizes the training-AUC on the current training set and add it to the current group of features

Update the current AUC with the new AUC obtained

When AUC==1 is reached, calculate the score composed from the current group of features on the test set and use it to obtain the test-AUC.

[00121] To limit the running time, the number of iterations (feature additions) of each internal round is limited by 15, but we note that in practice in all rounds an AUC of 1 on the training set was obtained before reaching this limit.

[00122] After 500 external round iterations, we assign a score that summarizes the predictive power of each feature f, score f = scoreC f - scoreE f ,

Where scoreC f is the number of successful iterations (iterations with test-AUC >=0.6) in which feature / has been selected to be in the prediction score set, and scoreE f is the number of unsuccessful iterations (iterations with test-AUC <=0.4) in which feature / has been selected. We select the features assigned with significantly high score f (with binomial p-value < 0.05), which result in a final set of 15 IMPRES features (out of 294 pairwise binary relations considered). Feature selection using all NB data samples

[00123] We repeat a similar feature selection process (described above in

'Normalization and preparation of the neuroblastoma data' for the cluster defined by PC2+PC3>0), analyzing all samples whose age<18 months and belonging to either the "spontaneous regression" or "no spontaneous regression" categories. The top selected features have only moderate performance on the NB data (Figure 4B), but 12 of the 18 features selected overlap with the IMPRES features (using similar threshold to that used for IMPRES of binomial P-value<0.05 for score f ) (Supp. Table 9). The prediction performance of the classifiers built from these features is shown in Figure 9.

Example 2:

[00124] To test the hypothesis that an immune-based predictor of NB spontaneous regression may effectively predict ICB response in melanoma, we built a predictor of spontaneous regression in NB, analyzing the transcriptomics data of 108 patients. Those include both spontaneously regressing (patients considered as low risk NB and with no tumor progression) and high risk progressing patients (i.e., without spontaneous regression,

Methods) 13. We focused on 28 immune checkpoint genes collected from the literature that were included in all RNA-sequencing (RNA-seq) datasets available to us (Supp. Table 1). We based the NB predictor on pairwise relations between the (normalized) expression levels of these genes. Each predictive feature compares the expression of two checkpoint genes A and B, capturing a logical relation between their transcriptional levels (e.g., A > B). We performed a feature selection procedure searching for a subset of these features that best separates spontaneously regressing NB patients from those with high risk progressing disease, resulting in 15 most predictive features (Methods). Based on these features, the prediction of spontaneous regression of a tumor sample from its expression data is simply made by counting the number of predictive feature pairs that are fulfilled (true) in that sample given its transcriptomics data. This number, ranging from 0-15, denotes its EVImuno- PREdictive Score (IMPRES), with higher scores predicting spontaneous regression (Methods; Supp. Table 2). The resulting predictor obtains an accuracy of 0.9 (in terms of the Area Under the Receiver Operator Curve (AUC)) in the NB dataset (Figure 4, Methods). Reassuringly, examining tumors derived from patients with melanoma who were not treated with ICB ¹⁴ , the IMPRES scores of patients denoted as 'high immune response' are considerably higher than that of other subtypes (Figure 1A). Additionally, we find that IMPRES is significantly and positively associated with higher overall survival in these datasets ¹⁴ (Figure IB).

[00125] We next turned to investigate whether there are similarities between the cellular processes mediating spontaneous tumor regression in NB and those mediating spontaneous and ICB -stimulated immune response in melanoma, studying 9 different melanoma datasets ^{1 6} . To this end we identified immune related, Consistently Differentially expressed Pathways (termed CDPs) in ICB responders versus non-responders (evaluated separately for patients treated with anti-PD-1 or anti-CTLA-4 treatments, Methods). We identified seven CDPs that are common across all anti-PD-1 datasets and four CDPs across all anti-CTLA-4 datasets. We find that these CDPs are also differentially expressed in a similar manner in the 'high immune response' melanomas compared with other subtypes ¹⁴ and in spontaneously regressing vs high risk progressing NB tumors (Figure 1C, Supp. Table 3) (Methods; Supp. Table 4 and Figure 5 for a related analysis based on estimated immune cell abundances). We then computed the correlations between each IMPRES feature (using expression ratios, Methods) and the expression of each of the CDPs. As evident from Figure ID, these associations are consistently maintained across the four sample groups studied.

[00126] We then turned to apply IMPRES to predict melanoma patients' response to

ICB, without any further training. To this end we analyzed 256 samples from 9 datasets derived from 6 independent studies including patients treated with anti-CTLA-4, anti-PDl or their combination ¹ ^. We computed the IMPRES score of each melanoma sample from its expression data and used those and the clinical response data to generate the Receiver Operator Characteristic (ROC) classification curves quantifying IMPRES prediction performance in each of the different datasets. The resulting AUCs are in the range 0.77-0.96 (Figure 2A).

[00127] We further tested the predictive ability of IMPRES in a newly generated

RNA-seq data of tumor biopsies from metastatic melanoma patients treated with ICB therapies at the Massachusetts General Hospital. IMPRES achieves AUCs of 0.81 and 0.97 on the anti-PD-1 and anti-CTLA-4 samples respectively (Figure 2B). It maintains its predictive accuracy when evaluating the aggregate collection of the datasets studied above (a total of 297 samples, Figure 2B). Figure 2C shows the number of true/false positives (responders) and true/false negatives (non-responders) obtained on this aggregated data at different IMPRES score classification thresholds, manifesting the well-known tradeoff between precision and recall (Figure 2D, Figure 6A-B, Supp. Table 5). As evident, IMPRES can capture almost all true responders while misclassifying less than half of the non- responders (at threshold = 8). Higher IMPRES scores are also associated with improved overall survival and progression-free survival (PFS) in ICB treated melanoma patients (Methods, Figure 2E-H, Figure 6C).

Supplementary Table 5. True positive and true negative rates for ICB response predictions in each melanoma dataset.

To compare the predictive accuracy of IMPRES with that of current transcriptome-based predictors, we generated predictors of response to ICB for each published transcriptomic signature (Methods). The performance of IMPRES is superior to the other predictors (Figure 3A, Supp. Table 6). This observation also holds true when we compare the performances on each ICB-treatment group separately (Figure 3B). Overall, the predictors built on biologically motivated scores (cytolytic-activity ¹⁵ and PDL-1 expression) generalize better than the machine learning based predictors constructed on transcriptomic signatures identified in the specific cohorts. Of note, while we find a significant correlation between IMPRES and CD8 ⁺ and CD4 ⁺ T cells abundances inferred via CIBERSORT, the latter are poor predictors of ICB response (Figure 7). IMPRES superiority is particularly notable because for most existing signature-based predictors (all but cytolytic-activity ¹⁵ and PDL-1 expression) we had to retrain the latter separately for each dataset, otherwise their overall performance was dismal, testifying to their poor generalizability between different datasets (Methods). In contrast, IMPRES is constructed only once from the NB data and never trained on any melanoma dataset. Thus, it is markedly less prone to over-fitting, a paramount concern regarding standard cancer transcriptomics predictors ^{16 18} . To further study the importance of training on the independent NB data, we trained ICB response predictors based on melanoma data instead of NB, following exactly the same representation and training procedure as used in IMPRES. This results in markedly lower prediction performances on the melanoma datasets that were not used for training compared to the original IMPRES procedure (Figure 8). Finally, IMPRES performance remains superior when repeating this comparative analysis while excluding patients annotated with 'stable disease' (Supp. Table 6).

[00128] The features composing IMPRES uncover a few insights that are biologically interesting. Reassuringly, the relatively higher expression of genes encoding immune stimulatory molecules (such as HVEM, CD27 and CD40) is associated with a better response to ICB, while the higher expression of genes encoding immune inhibitory molecules (such as CD276, TIM-3, CD200 and VISTA) is associated with a worse response, as expected (Figure 3C). Higher expression of CD40 compared to that of PD-1, PDL-1, CD80 and CD28 is associated with a better ICB response, in line with the recent findings that agonists of CD40 reverse resistance to anti-PD-1 therapy, and that induced PD-1 expression mediates acquired resistance to antagonist CD40 treatment ¹⁹ . Additionally, the higher expression of CD27 compared to that of PD1 (but not compared to CTLA-4) is associated with improved response. This is in line with recent findings that the combination of a CD27 agonist plus anti-PD- 1 recapitulates the effects of CD4+ T helper cells on tumor control, while the combination of a CD27 agonist plus anti-CTLA-4 did not improve tumor control 20.

[00129] We further studied the individual predictive power of the IMPRES features by considering the expression ratio of each predictive pair (Methods). We find that some features are specifically more predictive for anti-PD-1 pre-treatment (CD28/CD86, Rank-sum P-value = 0.05) or on-treatment (PD1/OX40L, CD86/OX40L and CD86/CD200, Rank-sum P-value = 0.018 for all, data not shown). Notably, no feature emerges as being strongly predictive of response to anti-CTLA-4 specifically (Figure 3D, and other data not shown). Examining the associations between these 15 features (using their expression ratio) and the inferred abundance of 22 types of immune cell (data not shown) uncovers two significant associations, with CD40/PD-1 and PD1/OX40L (Figure 3E). Finally, a feature reduction analysis (Methods) shows that the overall predictive performance of IMPRES can be maintained with a subset of 11 of the 15 original features, but beyond that it markedly decreases (Figure 9, Supp. Table 8).

[00130] In summary, IMPRES' high predictive performance is mainly due to two key conjectures: (a) key immune mechanisms underlining spontaneous regression in NB are shared with those determining response to ICB in melanoma, and (b) those may be captured by specific pairwise relations of immune checkpoint genes' expression. Building on these assumptions leads to a predictor of response to checkpoint therapy that is significantly superior to existing predictors and displays robust performance across many different melanoma datasets. From a translational standpoint, we show that IMPRES can correctly capture almost all true responders while misclassifying less than half of the non-responders.

References

1. Van Allen, E. M. et al. Genomic correlates of response to CTLA-4 blockade in metastatic melanoma. Science (80-. ). 350, 207-211 (2015).

2. Chen, P. L. et al. Analysis of immune signatures in longitudinal tumor samples yields insight into biomarkers of response and mechanisms of resistance to immune checkpoint blockade. Cancer Discov. 6, 827-837 (2016).

3. Hugo, W. et al. Genomic and Transcriptomic Features of Response to Anti-PD-1 Therapy in Metastatic Melanoma. Cell 165, 35-44 (2016).

4. Weinstein, J. N. et al. The Cancer Genome Atlas Pan-Cancer analysis project. Nat. Genet. 45, 1113-20 (2013).

5. Prat, A. et al. Immune-related gene expression profiling after PD-1 blockade in non- small cell carcinoma, head and neck squamous cell carcinoma, and melanoma. Cancer Res. 77, 3540-3550 (2017).

6. Riaz, N. et al. Tumor and Microenvironment Evolution during Immunotherapy with Nivolumab. Cell (2017). doi: 10.1016/j.cell.2017.09.028

7. Bramhall, R. J., Mahady, K. & Peach, A. H. S. Spontaneous regression of metastatic melanoma - Clinical evidence of the abscopal effect. Eur. J. Surg. Oncol. 40, 34-41 (2014).

8. Kalialis, L. V., Drzewiecki, K. T. & Klyver, H. Spontaneous regression of metastases from melanoma: review of the literature. Melanoma Res. 19, 275-282 (2009).

9. Diede, S. J. Spontaneous regression of metastatic cancer: learning from neuroblastoma. Nat. Rev. Cancer 14, 71-2 (2014).

10. Brodeur, G. M. & Bagatell, R. Mechanisms of neuroblastoma regression. Nat. Rev. Clin. Oncol. 11, 704-713 (2014).

11. Cheung, N. K. V. et al. Ganglioside G(D2) specific monoclonal antibody 3F8: A phase I study in patients with neuroblastoma and malignant melanoma. Journal of Clinical

Oncology 5, 1430-1440 (1987).

12. Yu, A. L. et al. Anti-GD2 Antibody with GM-CSF, Interleukin-2, and Isotretinoin for Neuroblastoma. N. Engl. J. Med. 363, 1324-1334 (2010).

13. Su, Z. et al. An investigation of biomarkers derived from legacy microarray data for their utility in the RNA-seq era. Genome Biol. 15, 523 (2014).

14. Jonsson, G. et al. Gene expression profiling-based identification of molecular subtypes in stage IV melanomas with different clinical outcome. Clin. Cancer Res. 16, 3356-3367 (2010).

15. Rooney, M. S., Shukla, S. A., Wu, C. J., Getz, G. & Hacohen, N. Molecular and genetic properties of tumors associated with local immune cytolytic activity. Cell 160, 48-61 (2015).

16. Simon, R., Radmacher, M. D., Dobbin, K. & McShane, L. M. Pitfalls in the use of DNA microarray data for diagnostic and prognostic classification. . Natl. Cancer Inst. 95, 14-18 (2003).

17. Tinker, A. V., Boussioutas, A. & Bowtell, D. D. L. The challenges of gene expression microarrays for the study of human cancer. Cancer Cell 9, 333-339 (2006).

18. Ransohoff, D. F. Bias as a threat to the validity of cancer molecular-marker research. Nat. Rev. Cancer 5, 142-149 (2005).

19. Zippelius, A., Schreiner, J., Herzig, P. & Muller, P. Induced PD-L1 Expression Mediates Acquired Resistance to Agonistic Anti-CD40 Treatment. Cancer Immunol. Res. 3, 236-244 (2015).

20. Ahrends, T. et al. CD27 Agonism Plus PD-1 Blockade Recapitulates CD4+ T-cell Help in Therapeutic Anticancer Vaccination. Cancer Res. 76, 2921-2931 (2016).

21. Chen, L. & Flies, D. B. Molecular mechanisms of T cell co- stimulation and co- inhibition. Nature Reviews Immunology 13, 227-242 (2013).

22. Zhang, Q. & Vignali, D. A. A. Co- stimulatory and Co-inhibitory Pathways in

Autoimmunity. Immunity 44, 1034-1051 (2016).

23. Fuertes Marraco, S. A., Neubert, N. J., Verdeil, G. & Speiser, D. E. Inhibitory receptors beyond T cell exhaustion. Frontiers in Immunology 6, (2015).

24. Ramsay, A. G. Immune checkpoint blockade immunotherapy to activate anti-tumour T-cell immunity. British Journal of Haematology 162, 313-325 (2013).

25. Buchbinder, E. I. & Desai, A. CTLA-4 and PD-1 Pathways: Similarities, Differences, and Implications of Their Inhibition. Am. J. Clin. Oncol. 39, 98-106 (2016).

26. Ashburner, M. et al. Gene ontology: Tool for the unification of biology. Nature Genetics 25, 25-29 (2000).

27. Newman, A. M. et al. Robust enumeration of cell subsets from tissue expression profiles. Nat. Methods 12, 453-457 (2015).

28. Jenkins, R. W. et al. Ex Vivo Profiling of PD-1 Blockade Using Organotypic Tumor Spheroids. Cancer Discov. CD-17-0833 (2017). doi: 10.1158/2159-8290.CD-17-0833

29. Eisenhauer, E. A. et al. New response evaluation criteria in solid tumours: Revised RECIST guideline (version 1.1). Eur. J. Cancer 45, 228-247 (2009).

30. Hoos, A., Wolchok, J. D., Humphrey, R. W. & Hodi, F. S. CCR 20th anniversary commentary: Immune-related response criteria - Capturing clinical activity in immuno- oncology. Clinical Cancer Research 21, 4989-4991 (2015).

31. Ayers, M. et al. IFN- γ - related mRNA profile predicts clinical response to PD-1 blockade. J. Clin. Invest. Ill, 1-11 (2017).

32. Gao, J. et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci. Signal. 6, pll (2013).