Field of the invention

The invention relates to the field of sample characterisation. More particularly, the present invention relates to methods and systems for analyzing samples using recorded UV-VIS spectra.

Background of the invention

Although numerous analysis techniques for qualification and quantification of samples exists, only few analysis techniques are as simple to perform, fast and accurate as spectrophotometry. One example of spectrophotometry is UV-VIS absorbance spectroscopy. During such experiments, samples are irradiated with UV- VIS radiation of different wavelengths, the radiation remaining after passage through the sample is detected and the absorbance at different wavelengths is determined. As particular components will show a particular absorbance at particular wavelengths, such a particular absorbance profile can be used as a fingerprint which allows, upon comparison with reference spectra, to identify the components. When more complex samples are studied, the absorbance features in the spectrum can be more broad, rendering the interpretation of spectra substantially more difficult.

A number of different spectral analysis techniques have been described in the past. Some methods are provided allowing determining the concentration or relative concentration of components of a sample based on the approximation of the measured spectrum with a linear combination of standard, also referred to as reference, spectra of the individual components. Approximation of the measured spectrum with reference spectra may be performed using minimization techniques for fitting the reference spectra to the absorption spectra, such as for example least squares regression or least absolute deviations regression. Such methods also may be iteratively applied.

More complex methods also are known, such as methods making use of derivatives of a spectrum, as amongst others the latter may allow to increase the insensitivity to wavelength shifts. One example of a method using derivatives is the analysis of a protein mixture using the second derivative spectrum of the sample. Another known method is based on matrix computation for determining the concentration of a number of chemical components using a number of predetermined spectral data of each chemical component, whereby the number of predetermined spectral data is larger than the number of chemical components to be determined. Some of the known methods are based on cross-correlation of samples and reference spectra, both being weighted by mean transmittance per wavelength. Particular applications for analyzing samples with known composition have been described, such as for example analysis of the composition of a protein using electromagnetic spectroscopy, analysis of RNA wherein samples are chemically split in AC and GU fractions and the AC and GU composition is then determined based on absorption at 260nm.

Furthermore, also some methods are known that deal with unknown component spectra. Methods are known wherein the separation of unknown component spectra from samples is based on constrained linear or quadratic programming and sample transforms, i.e. wavelet transforms. Other methods make use of approximated component spectra by point-wise cross-correlation of absorbance and making use of known concentration. Several techniques make use of principal component analysis (PCA) for obtaining diagnostically relevant reference spectra from a database of spectra associated with disease states whereby correlation with reference spectra is used to aid in the diagnosis of a new sample.

Spjotvoll et al. describe in Technometrics 24 (1982) a technique for determining spectra and concentrations of two-constituent mixtures based on a least squares estimator. A method is discussed for identification of two unknown chemical compounds and an estimation of their proportion in a set of unknown mixtures of the two compounds. It is based on a least-squares fit and principle component analysis for separation whereby additional constraints are introduced such as the sum of concentrations being equal to unity and spectra and concentrations being non- negative.

UV-VIS spectral analysis typically is complicated significantly when the samples that are studied suffer from turbidity. Turbidity is the effect of cloudiness or haziness of a fluid caused by suspensed particles that are individually generally invisible to the naked eye. The occurrence of turbidity often results in samples not being able to be characterized, or samples needing to be processed prior to spectroscopic measurement, in order to reduce or remove turbidity.

There still is a need for a method allowing good characterisation of complex mixtures comprising more than two components, particularly using UV-VIS spectroscopy.

Summary of the invention

It is an object of embodiments of the present invention to provide good methods and systems for UV-VIS spectroscopic analysis of samples suffering from turbidity. The UV-VIS spectroscopic analysis may be UV-VIS absorbance spectroscopic analysis. It has surprisingly been found that, rather than requiring pre-processing of samples to be studied for removing turbidity and/or scatter, the contribution of turbidity and/or scatter in the obtained UV-VIS spectroscopic data can be significantly accurately determined in data analysis so that further characterisation of the sample can be obtained based on such UV-VIS spectroscopic data recorded for samples suffering from turbidity and/or scatter.

It is an advantage of embodiments of the present invention that a powerfull data analysis of UV-VIS spectroscopy spectra is obtained.

It is an advantage of embodiments according to the present invention that good quantification can be obtained, both for routine applications and more complex applications.

It is an advantage of embodiments according to the present invention that based on prior knowledge in one or more samples, accurate reference spectra can be obtained for components in the sample.

It is an advantage of embodiments according to the present invention that based on prior knowledge of reference spectra of one or more components in the one or more samples, accurate concentration information can be obtained for components in the sample.

It is an advantage that methods and systems according to embodiments of the present invention allow for obtaining relevant and conclusive results. The above objective is accomplished by a method and device according to the present invention.

The present invention relates to a method for analyzing UV-VIS spectrophotometer data of one or more samples , the one or more samples consisting of a number of constituents, the method comprising obtaining prior information for the one or more samples regarding their constituents, obtaining UV-VIS spectrophotometer data for the one or more samples, defining a number of overlapping components contributing in the UV-VIS spectrophotometer data, the number of overlapping components comprising one or more components assigned to known constituents of the one or more samples and the number of components comprising one or more components that cannot be assigned to known constituents of the one or more samples, using the prior information for the one or more samples regarding their constituents and using the UV-VIS spectrophotometer data, estimating the constituents composition and the component contributions to the UV-VIS spectrophotometer data for the number of components for the one or more samples by minimizing a residue between the UV- VIS spectrophotometer data and a fit based on said constituent composition and said component contributions, thus obtaining information regarding the one or more components assigned to known constituents of the one or more samples and regarding the one or more components that cannot be assigned to known constituents of the one or more samples, wherein estimating the constituents and the component contributions comprises one or more samples including one or more components being representative for turbidity caused by concentration or aggregation effects and/or for scattering caused by residual particles . It is an advantage of embodiments according to the present invention that the analysis method allows for accurate detection of contributions to the spectrum that stem from modified constituents, interaction of different constituents or contaminations in the one or more samples.

Said taking into account components may comprise correcting the UV-VIS spectrophotometer data by fitting with one or more correction spectral components representative for turbidity caused by concentration or aggregation effects and/or for scatter caused by residual particles.

Said estimating may comprise correcting the UV-VIS spectrophotometer data by fitting in a predetermined wavelength range only and deriving a functional shape of the one or more correction spectral components based thereon.

The functional shape of the one or more correction spectral components may be a power function of the inverse of the wavelength or a linear combination of power functions of the inverse of the wavelength.

Said correcting may be performed prior to said minimizing.

Said correcting may be performed together with said minimizing.

Estimating the constituents composition and the component contributions to the UV- VIS spectrophotometer data for the number of components for the one or more samples by minimizing a residue between the UV-VIS spectrophotometer data and a fit based on said constituent composition and said component contributions may be performed iteratively until a residue smaller than a predetermined level is obtained. It is an advantage of embodiments according to the present invention that using an iterative method an increased accuracy for the fit and the corresponding constituent composition and component contribution can be obtained.

The method furthermore may comprise cross-checking the prior information based on any or more of the constituents composition, the component contribution or the residue between the UV-VIS spectrophotometer data and the fit. It is an advantage of embodiments that a cross-check regarding the obtained results can be performed using the prior information regarding the one or more samples.

Said prior information may comprise reference contributions in UV-VIS spectrophotometer data for components assigned to known constituents and said estimating the constituents composition and the component contributions may comprise estimating the component contributions to the UV-VIS spectrophotometer for all components based on the reference contributions and determining an estimate for the constituents composition based on minimization of the residue between the UV-VIS spectrophotometer and the fit based on said estimated component contributions to the UV-VIS spectrophotometer data for all components. It is an advantage of embodiments according to the present invention that accurate composition information of the one or more samples can be obtained based on reference spectra of constituents present in the one or more samples, taking into account the presence of unknown contributions in the UV-VIS spectrophotometer data.

Said prior information may comprise constituent composition information for the one or more samples for the known constituents and said estimating the constituents composition and the component contributions may comprise estimating the constituent composition for all constituents based on the prior information on the constituent composition information for the one or more samples for the known constituents, and determining an estimate for the component contribution to the UV- VIS spectrophotometer data based on minimization of the residue between the UV- VIS spectrophotometer data and the fit based on said estimated constituent composition for all constituents. It is an advantage of embodiments according to the present invention that accurate composition information of the one or more samples can be obtained based on reference spectra of constituents present in the one or more samples, taking into account the presence of unknown contributions in the UV- VIS spectrophotometer data.

The constituent composition and the component contributions may be interrelated by a model defining weighing factors for component contributions as function of the constituent composition and wherein determining an estimate for the constituent composition or for the component contribution comprises taking into account the weighing factors. It is an advantage of embodiments according to the present invention that the composition of constituents in the samples can be taken into account.

Analysis may be performed on spectrophotometer data for a plurality of samples, and said iteratively estimating may be performed for those samples having the smallest residue. The components that cannot be assigned to known constituents of the one or more samples may be representative for any or more of modified constituents, neighbouring effects of different constituents or contaminations in the one or more samples. The method may comprise outputting a contribution related to turbidity and/or scattering.

It is an advantage of embodiments of the present invention that good and/or accurate quantification of components in samples showing high turbidity can be obtained, e.g. based on prior knowledge of reference spectra of components in the sample. Another component that advantageously can be taken into account by components that cannot be assigned to known constituents of the one or more samples may be representative for scatter by residual microparticles. Such particles may be non-intentionally present in the sample, such as for example introduced by earlier processing of the sample. In one example, such residual particles may be magnetical micro-particles. The method may comprise outputting a contribution related to such scattering from residual particles. It is an advantage of embodiments of the present invention that good and/or accurate quantification of components in samples comprising residual particles can be obtained, e.g. based on prior knowledge of reference spectra of components in the sample.

The method may comprise using spectrophotometric data of a plurality of samples, wherein the samples comprises at least one sample wherein at least one unknown constituent is present and at least one sample wherein the at least one unknown constituent is absent.

The constituents may comprise a plurality of oligo's and the prior information may comprise an oligonucleotide present in the sample.

The number of components may be determined by component analysis.

The number of components may be determined based on principal component analysis.

The present invention also relates to a system for analyzing UV-VIS spectrophotometer data of one or more samples , the one or more samples consisting of a number of constituents, the system comprising an input means for obtaining prior information for the one or more samples regarding their constituents and for obtaining UV-VIS spectrophotometer data for the one or more samples, a processing means programmed for defining a number of components contributing in the UV-VIS spectrophotometer data, the number of components comprising one or more components assigned to known constituents of the one or more samples and the number of components comprising one or more components that cannot be assigned to known constituents of the one or more samples, the processing means furthermore being programmed for using the prior information for the one or more samples regarding their constituents and using the UV-VIS spectrophotometer data, estimating the constituents composition and the component contributions to the UV- VIS spectrophotometer data for the number of components for the one or more samples by minimizing a residue between the UV-VIS spectrophotometer data and a fit based on said constituent composition and said component contributions, thus obtaining information regarding the one or more components assigned to known constituents of the one or more samples and regarding the one or more components that cannot be assigned to known constituents of the one or more samples. The processing means is programmed for taking into account components that cannot be assigned to known constituents of the one or more samples including one or more components being representative for turbidity caused by concentration or aggregation effects and/or for scatter caused by residual particles.

The present invention also relates to the use of a method as described above for identifying and/or quantifying contaminants in DNA or RNA samples. Said contaminants may be protein contamination. The contaminants may be PCR- inhibiting contaminants.

The present invention also relates to the use of a method as described above for quantifying an amount of double-stranded DNA in a mixture of double-stranded DNA and single-stranded RNA or DNA.

The present invention also relates to the use of a method as described above for obtaining the composition of a polynucleotide or protein. The present invention also relates to the use of a method as described above for quantification of the modification efficiency in a sample.

The present invention also relates to the use of a method as described above for validating a sample composition or constituent thereof.

The present invention also relates to a computer program product for, if implemented on a processing unit, performing a method as described above. It also relates to a data carrier storing the computer program product and transmission of the computer program product over a network.

Particular and preferred aspects of the invention are set out in the accompanying independent and dependent claims. Features from the dependent claims may be combined with features of the independent claims and with features of other dependent claims as appropriate and not merely as explicitly set out in the claims. These and other aspects of the invention will be apparent from and elucidated with reference to the embodiment(s) described hereinafter. Brief description of the drawings

FIG. 1 illustrates a flow chart of a method for analyzing spectra according to an embodiment of the present invention.

FIG. 2 illustrates an example of data flow as can be obtained using embodiments of the present invention.

FIG. 3 illustrates an example of data flow as can be obtained using embodiments of the present invention.

FIG. 4 and FIG. 5 illustrate examples of UV-VIS spectroscopy data for a set of samples recorded with (FIG. 5) and without (FIG. 4) centrifuging, illustrating aspects according to embodiments of the present invention.

FIG. 6 and FIG. 7 illustrate deconvoluted spectra of a same DNA sample recorded with (FIG. 7) and without (FIG. 6) centrifuging obtained using a method for deconvoluting according to embodiments of the present invention.

The drawings are only schematic and are non-limiting. In the drawings, the size of some of the elements may be exaggerated and not drawn on scale for illustrative purposes. Any reference signs in the claims shall not be construed as limiting the scope.

In the different drawings, the same reference signs refer to the same or analogous elements.

Detailed description of illustrative embodiments

The present invention will be described with respect to particular embodiments and with reference to certain drawings but the invention is not limited thereto but only by the claims. The drawings described are only schematic and are non-limiting. In the drawings, the size of some of the elements may be exaggerated and not drawn on scale for illustrative purposes. The dimensions and the relative dimensions do not correspond to actual reductions to practice of the invention.

Furthermore, the terms first, second and the like in the description and in the claims, are used for distinguishing between similar elements and not necessarily for describing a sequence, either temporally, spatially, in ranking or in any other manner. It is to be understood that the terms so used are interchangeable under appropriate circumstances and that the embodiments of the invention described herein are capable of operation in other sequences than described or illustrated herein.

Moreover, the terms top, under and the like in the description and the claims are used for descriptive purposes and not necessarily for describing relative positions. It is to be understood that the terms so used are interchangeable under appropriate circumstances and that the embodiments of the invention described herein are capable of operation in other orientations than described or illustrated herein.

It is to be noticed that the term "comprising", used in the claims, should not be interpreted as being restricted to the means listed thereafter; it does not exclude other elements or steps. It is thus to be interpreted as specifying the presence of the stated features, integers, steps or components as referred to, but does not preclude the presence or addition of one or more other features, integers, steps or components, or groups thereof. Thus, the scope of the expression "a device comprising means A and B" should not be limited to devices consisting only of components A and B. It means that with respect to the present invention, the only relevant components of the device are A and B. Reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment, but may. Furthermore, the particular features, structures or characteristics may be combined in any suitable manner, as would be apparent to one of ordinary skill in the art from this disclosure, in one or more embodiments.

Similarly it should be appreciated that in the description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. This method of disclosure, however, is not to be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.

Furthermore, while some embodiments described herein include some but not other features included in other embodiments, combinations of features of different embodiments are meant to be within the scope of the invention, and form different embodiments, as would be understood by those in the art. For example, in the following claims, any of the claimed embodiments can be used in any combination.

Where in embodiments of the present invention reference is made to spectrophotometer data, reference is made to a reflectance, transmittance or absorbance intensity at a given wavelength or as function of wavelength. Advantageously, the method can be applied to spectrophotometer data being a spectrophotometer spectrum, indicating an intensity as function of wavelength. In advantageous embodiments, the data are absorbance data, representative of the absorption that has occurred in the one or more samples.

Embodiments of the present invention are directed to methods and systems for analyzing UV-VIS spectrophotometer data. Where reference is made to UV-VIS, reference is made to a wavelength region having an upper wavelength in the region 700nm to llOOnm and a lower wavelength in the region 150nm to 300nm.

Where in embodiments of the present invention reference is made to components contributing to the UV-VIS spectrophotometer data, reference is made to components to an intensity fraction for one or more wavelengths in the spectrophotometer data, e.g. spectrophotometer spectrum. Such a component may be directly assignable to a constituent in the one or more samples or may be not or not directly assignable to a constituent in the one or more samples.

Where in embodiments of the present invention reference is made to components that cannot be assigned to known constituents of the one or more samples, reference is made to components that do not correspond with or for which it is not yet known that these correspond with a physical-chemical absorption characteristic of the constituents of the one or more samples. The components thus do not necessarily correspond to a physical-chemical absorption characteristic of the constituents. The components do not need to correspond with constituents having a significant contribution, e.g. absorbance, within the wavelength region under study.

Where in embodiments of the present invention reference is made to constituents of one or more samples, reference is made to biological or non- biological elements such as chemical elements that are present in the sample. Such constituents can for example be one of the following or elements or groups of elements thereof : biopolymers, small organic molecules, metabolites such as glucose or ethanol, proteins, peptides, nucleic acid segments, oligonucleotides, oligonucleotides containing one or more modifications such as for example fluorescent dye labels or chromophores, chemical linker modifications such as aldehyde, photphate, amine- or thiol-modifications, affinity-labels such as biotin, reactive components such as enzymes (ALP, HRP), alternative bases such as deoxyuracil, deoxyinosine, phosphothiolates, locked nucleic acids, polynucleotides, nucleotides, oligos, peptide nucleic acids, antibodies and variations thereof such as for example nanobodies or alphabodies, micro- and nanoparticles either intentionally or unintentionally added, magnetic or non-magnetic particles, lipids, micelles, vesicles, viral particles, metal-complexes, interacallating dyes and polymers such as polythiophene-derivatives, chromogens such as p-NPP, tetrazolium salts, staining reagents such as Coomassie, methyl blue, eosin, molecules such as small molecule pharmaceuticals, antibiotics or drugs, molecules with a regulatory effect in enzymatic processes such as promoters, activators, inhibitors, or cofactors, viruses, bacteria, cells, cell components, cell membranes, spores, DNA, RNA, micro-organisms and fragments and products. Such constituents also can be for example chemical elements, such as chemical elements present in a pharmaceutical or chemical product.

Where in embodiments according to the present invention reference is made to overlapping components, reference is made to the fact that the component contribution to the spectrum of such components is significant at the same wavelength or in the same wavelength range.

Where in embodiments of the present invention reference is made to turbidity, reference is made to cloudiness or haziness of a fluid caused by individual particles that are generally invisible to the naked eye.

In a first aspect, the present invention relates to a method for analyzing UV-VIS spectrophotometer data of one or more samples. A typical disadvantage of UV-VIS spectrophotometer data of complex data is that different contributions e.g. corresponding with different constituents in the sample or with unknown constituents, will result in overlapping contributions. This results in broad-banded spectra, whereby the different contributions are not as such identifiable from the spectra in the form of shoulders or peaks. Embodiments of the present invention allows analyzing of UV-VIS spectrophotometer data of such complex samples resulting in spectra with broad band absorption built up by overlapping contributions. Methods according to embodiments of the present invention can be used for analysis of one or more samples, especially samples wherein turbidity and/or scatter plays a significant role. Whereas analysis of a single sample is possible, the method is especially suitable for analyzing spectrophotometer data of multiple samples. The one or more samples comprise a number of constituents. The methods according to embodiments of the present invention are especially suitable for analyzing data for samples having two or more, advantageously three or more constituents, whereby one of the components is representative for turbidity and/or scatter. By way of illustration, an exemplary flow chart of a method according to an embodiment of the present invention, comprising standard and optional steps is shown in FIG. 1, and will be further described with reference thereto below.

The method 100 according to embodiments of the present invention comprises obtaining 120 prior information for the one or more samples regarding their constituents. Such prior information for the one or more samples regarding their constituents may comprise in some embodiments one or more reference spectra. Such prior information alternatively or in addition thereto also may comprise expected composition information, such as for example expected concentrations, expected ratios between different constituents, etc. One example of such composition information may for example be the presence of a known polynucleotide, whereby the known polynucleotide is a known sequence of nucleotides such that known ratios of constituents, in the present example being nucleotides, are available. In some embodiments the prior information may be obtained based on the samples under study. In some embodiments the prior information may be previously stored information, such as for example previously stored reference spectra for constituents. In some embodiments a combination of such prior information also can be used.

The method also comprises obtaining 110 UV-VIS spectrophotometer data for the one or more samples. Such UV-VIS spectrophotometer data typically may be a spectrophotometer spectrum, although information at one or more individual wavelengths or wavelength ranges also may be used. Advantageously, the spectrophotometer data comprises a spectrum comprising at least a subrange of the wavelength range 600nm to 300nm, advantageously in the range 350nm to 550nm, or e.g. a subrange having a lower endpoint in the range 300nm to 359nm and a higher endpoint in the range 450nm to 600nm or higher. Alternatively or in addition thereto, specific absorbance or reflectance values at one or more wavelengths in the range 300nm to 600nm, advantageously in the range 350nm to 550nm may be part of the spectrophotometer data obtained. In advantageous embodiments, spectrophotometer data for wavelengths where no contribution or no strong contribution of specific constituents of the sample occurs are included, as these advantageously can be used for fitting one or more component representative for turbidity and/or scatter.

The method according to the present invention also comprises defining 130 a number of overlapping components contributing in the UV-VIS spectrophotometer data, the number of overlapping components comprising one or more components assigned to known constituents of the one or more samples and the number of components comprising one or more components that cannot be assigned to known constituents of the one or more samples. The components according to embodiments of the present invention expressing the contribution to the UV-VIS spectrophotometer data may in some embodiments comprise a spectral contribution with a shape differing from a Gaussian, Lorentzian or a mixture thereof. The shape of the spectral contribution may be a more broad contribution. At least one of these components that cannot be assigned to known constituents of the one or more samples is, according to embodiments of the present invention, representative for turbidity and/or scatter. Such turbidity typically is caused by concentration or aggregation effects and scatter is typically caused by residual particles. Further, one or more components that cannot be assigned to known constituents of the one or more samples may be as described above, such as for example modified constituents, neighboring effects between two constituents or contaminants. Defining the number of overlapping components may be based on a predetermined algorithm, a neural network, etc. The method may make use of a component analysis for defining the number of overlapping components, such as for example using a principal component analysis technique. In some embodiments the number of independent components can be obtained by putting a large number of absorbance spectra of the sample into a matrix and calculating a rank of the matrix.

The method according to the present invention furthermore comprises, using the prior information for the one or more samples regarding their constituents and using the UV-VIS spectrophotometer data, estimating 140 the constituents composition and the component contributions to the UV-VIS spectrophotometer data for the number of components for the one or more samples by minimizing a residue between the UV-VIS spectrophotometer data and a fit based on said constituent composition and said component contributions. According to embodiments of the present invention, for this estimation, the method comprises taking into account the one or more components that cannot be assigned to known constituents and that are representative for turbidity and/or scatter. The latter can be done in a variety of ways. In one embodiment, first a fitting of the one or more components that cannot be assigned to known constituents and that are representative for turbidity and/or scatter is performed to the spectrophotometer data, e.g. in a particular wavelength range, resulting in intermediate spectrometer data. The intermediate spectrometer data than is further deconvoluted in its further components. In other words, in one embodiment, a two step fitting process is performed. Alternatively, also a different number of fitting steps may be applied, such as for example a one-step fitting. Advantageously, fitting of the one or more turbidity and/or scatter related components is performed earliest in the fitting procedure, i.e. during the first step of the fitting, or if only a single fitting step is used, in the single fitting step. For minimizing a residue different techniques may be employed such as for example a least square method, iterative error minimization, error entropy minimization, weighted error minimization,...

In one example, the fitting can be performed in a region where a substantial contribution of another component is present. In such a case, a combination of the turbidity component and the other contributing component can be fitted to a region of the spectrum and the contribution (quantity) of the turbidity component can be determined. This turbidity contribution can then be taken into account when further deconvoluting the spectrum. The contribution of the other contributing component, determined while fitting the turbidity contribution, can either be used in the further deconvolution of the spectrum or can be reset and determined together with the other components contributing to the spectrum.

As indicated above, fitting of the one or more turbidity and/or scatter related components may be performed in a limited wavelength range of the spectrometer data according to a predetermined functional behavior, whereby the same functional behavior is used for the full wavelength range of the spectrometer data. Advantageously, fitting to the spectrometer data is performed in a limited wavelength range where no contribution or no substantial contributions of other components is expected. In some examples, the fitting may advantageously be performed in a subrange in the wavelength range 600nm to 300nm, advantageously in the range 350nm to 550nm, or e.g. a subrange having a lower endpoint in the range 300nm to 350nm and a higher endpoint in the range 550nm to 600nm or higher. Alternatively or in addition thereto, the fitting may also be performed at one or more particular wavelengths in the range 300nm to 600nm, advantageously in the range 350nm to 550nm.

According to embodiments of the present invention, the functional behavior used for fitting may be a predetermined functional behavior. The functional behavior for the one or more components may for example be a function proportional to the n-th power of the inverse of the wavelength. In case more components representative for turbidity and/or scatter are used, this may result in a linear combination of two functions being proportional to the n-th power of the inverse of the wavelength, i.e. a powerfunction of the inverse of the wavelength.

In one example, one or more components representative for turbidity and/or scatter thus have a shape according to the following functional dependency : component _turbidity oc

whereby component _tU rbidity is a first component being representative for turbidity and/or scatter, λ is the wavelength, and n is a real number between 1.5 and 8, e.g. between 1 and 4.

In another example the total turbidity correction may comprise more components and have a shape being a linear combination of powerfunctions of the inverse of the wavelength, for example for a linear combination of two powerfunctions resultin in component _{totalturbidity} = component^^ + component^^

with com ponenttotalturbidity being the overall turbidity and/or scatter contribution expressed in several correction components relating to turbidity and/or scatter componentturbidityi and component _tU rbidity2, wherein Ai and A ₂ are constants, λ is the wavelength, and χ\ and n ₂ are real numbers between 1.5 and 8. In one advantageous example, 1.8 and 4 are used for ni and n ₂ .

In some embodiments, the prior information comprises reference contributions in the UV-VIS spectrophotometer data, such as for example reference spectra, for components assigned to known constituents. Estimating the constituents composition and the component contributions then may comprise estimating the component contributions to the UV-VIS spectrophotometer data for all components based on the reference contributions. Based on the component contributions, an estimate of the constituent composition is then determined based on the minimization of the residue between the UV-VIS spectrophotometer and the fit based on the estimated component contributions to the UV-VIS spectrophotometer data for all components, thereby thus either taking into account turbidity and/or scatter in an, e.g. first, sub-step of the fitting procedure or during a single-step fitting procedure. In some embodiments, the prior information comprises constituent composition information for the one or more samples for the known constituents. Estimating the constituents composition and the component contributions then comprises estimating the constituent composition for all constituents based on the prior information on the constituent composition information for the one or more samples for the known constituents, and determining an estimate for the component contribution to the UV-VIS spectrophotometer data based on minimization of the residue between the UV-VIS spectrophotometer data and the fit based on said estimated constituent composition for all constituents.

In some embodiments a mixture of prior information is obtained, whereby the mixture of information allows for determining an estimate of the constituents composition and the component contributions.

The constituent composition and the component contributions may be interrelated by a model defining weighing factors for component distributions as function of the constituent composition. Such weighing factors may be taken into account when determining an estimate for the constituent composition or the component contribution.

The method also may comprise repeating 150 the estimating step and fitting step for further minimizing the total residue by iteratively applying these steps. Such iteration process may be performed until the remaining residue between the UV-VIS spectrophotometer data and the fit based on the constituent composition is smaller than a predetermined value, or until a maximum number of iteration steps would be reached. In one embodiment, fitting the components related to turbidity and/or scatter may be performed a single time, while fitting the further components may be performed iteratively. Alternatively, also the turbidity and/or scatter components fitting step may comprise a number of iterations. The predetermined value referred to may be based on predetermined rules, based on a neural network, based on predetermined algorithms, based on information regarding the one or more samples, etc. In some embodiments, minimization of the residue may only be performed for those samples that have the smallest residue, allowing obtaining more accurate results and/or quicker convergence.

The method allows providing as an output information regarding the composition of the sample and/or information regarding the spectral contribution of constituents to the absorbance spectra. In some examples, the obtained information may allow deriving a reference spectrum for a constituent, a modified constituent, a contaminant, a neighbouring effect, ... The latter can be used for setting up a library of different reference spectra compatible with different constituents or effects thereof.

In one embodiments, the method also comprises using the residue of each sample or information related thereto to check the available a priori information. In other words, the results obtained using the method can be used for cross-checking the available a priori information.

The method according to embodiments of the present invention may especially be suitable for analyzing a plurality of samples, whereby in one part of the samples the unknown component is present and in another part of the samples the unknown component is absent.

The method according to embodiments of the present invention may be implemented as software as well as as hardware.

By way of illustration, embodiments of the present invention not being limited thereto, an example of how data can be combined and obtained is illustrated in FIG. 2. FIG. 2 illustrates how, according to embodiments of the present invention, based on spectrophotometer data and composition information (e.g. obtained based on prior knowledge), component contributions can be determined, e.g. in the shape of reference spectra by minimizing the residue, taking into account at least one component representative for turbidity and/or scatter. Based on the obtained component contributions and using the spectrophotometer data, improved estimates of the composition can be obtained. FIG. 3 illustrates a further embodiment of the present invention, whereby upon estimation of the composition information or the component contribution information, it can be decided to add an additional component or remove a component by evaluating the component contribution data obtained e.g. in the light of the expected results in view of the prior data, or it can be decided to remove a sample in the evaluation based on an evaluation of the composition information obtained e.g. in the light of the expected results in view of the prior data. Again at least one component being representative for turbidity and/or scatter is taken into account, at least initially. The method may be suitable for extracting quantitative information about one or more constituents in the sample and the corresponding component contribution in the UV-VIS spectrum. In some embodiments, corresponding reference spectra of known and unknown components suspected to the present in the sample, can be used for fitting the UV-VIS spectrophotometer data. The component contributions may not be limited to the reference spectra of the individual pure components, but may be extended with corrective spectral functions of components that do not necessarily correspond to a physic-chemical absorption characteristic of the component, nor does the component itself needs to show significant absorption characteristics within the wavelength region under study. In examples according to embodiments of the present invention, at least one of the component contributions is a contribution corresponding with turbidity caused by concentration or aggregation effects. In one example, one of the component contributions may be a contribution corresponding with scatter caused by residual particles, e.g. particles present due to pre-processing of the sample.

By way of illustration, embodiments of the present invention not being limited thereto, an algorithm comprising a number of particular steps will be described illustrating an exemplary way of implementing the method according to embodiments of the present invention. The method will be described using a particular mathematical matrix formalism, although embodiments of the present invention are not limited thereto another mathematical formalism also may be used. The example illustrates the methods used for deconvolution of spectrophotometric data. The illustrative algorithm detailed below can be applied for a variety of samples. First, for the ease of explanation, the situation will be discussed whereby no turbidity components are to be taken into account, allowing to illustrate the principles of the algorithm used.

For illustrating the algorithm, a number of definitions first are provided : : this is a measurement (expressed in absorbances, ODIOm m) where the subscript i denotes the wavelengths (ranging from 220nm to 360nm in lnm steps, in case of nucleic acids)

Q _j : this matrix contains the relative presences of the different known and unknown constituents or contaminant of the mixture, j is the number of subcomponents in the mixture.

R. : the reference spectra. These are the (known) spectra of the known components in the sample. With i the wavelengths and j the subcomponents.

When working with stored reference spectra, R _j , the coefficients of each reference spectrum, Q _tj , can be the unknown. These typically are a measure for the amount of the component being present.

The measurements, relative presences and contributions (in the present example being known reference spectra) are given as

being a mathematical expression of the fitting procedure.

The definitions in case a plurality of samples are measured result in the measurement matrix becoming

Α : this is a measurement (expressed in absorbances, ODIOmm) where the subscript i denotes the wavelengths (ranging from 220nm to 360nm in lnm steps, in case of nucleic acids) and k the number of the measurements.

The coefficient matrix becoming

Q _jk : this matrix contains the - typically unknown - coefficients (the relative presences) of every subcom ponent of the mixture. Again j is the number of subcomponents, and k the number of measurements.

Resulting in the following equation expressing the relation between the components (in the present example known from prior info and being reference spectra R) the coefficient matrix (in the present example expressing the constituent composition) and the measurement matrix. RnQ _Jk = _k or R Q [2]

In the present example, the available prior information corresponds with the reference spectra for the constituents in the sample. In such a case both the components contribution information R and the spectrophotometer data A in equation [2] are known. To determine the sample composition Q first left-multiply the equation with the component contribution information R transposed and then left-multiply both sides of the equation with the inverse of the matrix product of the component contribution R transposed and the component contribution R

This is in fact the equation for the matrix linear regression or linear least squares. The expression in the right side before the spectrophotometer data A is known as the pseudo-inverse of the matrix R and can be for example calculated using singular value decomposition. In the example discussed above for ease of understanding, the situation was shown for known constituents. In embodiments according to the present invention, at least one component corresponds with an unknown constituent, more particularly a component representative for turbidity and/or scatter. In the following, it will thus be considered that the component contributions comprise not only contributions in the absorbance spectra of known species, but also for example contributions due to environmental interaction, the presence of modifications and the presence of contaminants whereby at least one component is representative for turbidity and/or scatter. Below, deconvolution of a spectrum and determination of the component contributions is targeted.

The component contribution matrix can be expressed as follows, whereby not only the component contributions B are incorporated - B being representative for all possible components including contaminant contributions C are taken into account, the contaminants comprising at least one component C _tU rbidity being representative for turbidity.

Q ^T = [B | C]

The latter case is an illustration of an algorithm as can be used for embodiments of the present invention, whereby at least one of the contaminants is representative for turbidity or scatter. As the contributions may not directly express the sample composition in terms of concentrations, the matrix Q can also be written as.

Q ^T =[B.d I C.C ₂ ]

Here the dot between the matrices is not to be interpreted as the matrix product but as an entrywise product (or Hadamard product) with :

CI : the sample component concentration matrix, a column vector with the concentration of each component. Every number of a row of B is multiplied by the number of Ci on the same row

C2: the contaminant concentration matrix, a rectangular matrix with the same dimension as C. This matrix contains the concentration of each contaminant in each sample as a relative number. A higher number corresponds to a higher concentration of the given contaminant in the sample. According to embodiments of the present invention, at least one element in the contaminant concentration matrix is representative for turbidity and/or scatter.

I n an analoguous way, matrix R can be split up in 2 matrices Rl and R2 (R = [Rl | R2]): Rl: component spectral contributions of the known components, corresponding to the first n columns of R, with n = number of columns of [B] . These component contributions may, depending on the prior information, be available as reference spectra.

R2: component contributions representative for the contaminants. I n some examples, as illustrated above, one or more of the component contributions representative for turbidity and/or scatter may show a functional dependency being proportional to a power function of the inverse of the wavelength.

According to at least some embodiments of the present invention, the spectral analysis may be performed such that first the component contributions representative for turbidity and/or scatter are determined. The latter may for example be performed by a residue minimization process for a particular wavelength region of the spectrum using predetermined functions in which no other contributions occur. The spectral contributions may be fitted for example in one illustration as power functions of the inverse of the wavelength.

The spectral contributions for the components representative for turbidity and/or scatter co example be calculated as :

I turbidity

whereby C _tur bidity comprises the contribution expressing the amount of turbidity and/or scatter present, being a portion of [C.C ₂ ]

and tur idty is a reference spectrum expressing the functional shape of the turbidity and/or scatter contribution, being a portion of [R ₂ ] . R _tU rbidity may for example be a power function of the inverse of the wavelength.

For determining the turbidity contribution Qurbidity a comparison typically is made to regions of the measured spectra where no other contributions are expected. The contributions are calculated by fitting, e.g. based on minimalisation processes and an iterative fit procedure. Alternatively, also a region of the measured spectrum can be used where a known component contributes, and this known component can also be fitted, using e.g. a reference spectrum for the known component. Again, as a result, the turbidity contribution Qurbidity can be determined.

In one example, the spectral contributions of turbidity and/or scatter can be determined separately using such a fitting procedure. After the contribution of the turbidity and/or scatter is determined, the spectrum can be corrected resulting in intermediate spectral data A', which can be used for performing the fit for determining the further deconvolution of the spectrum.

A'= A - R turbidity turbidity

The intermediate spectral data A' can then thus be used for determining the further deconvolution of the spectrum, using the techniques as described above.

Alternatively, the fitting of the turbidity component can also be combined with the fitting of the other components in a single fitting procedure. Using the reference spectra of the known components, calculate a predicted spectrum based on the reference spectra R and a first estimate of the component contributions Q. Such values may be initiated at a standard value. The component contributions Q can then be further determined by minimizing the difference between the intermediate spectral data and the predicted spectrum, which typically is done in a minimization procedure. Therefore,

a residue calculation can be performed, in the present example using the following steps :

· calculate the fitted spectra by multiplying the matrices Q and R

• Compare these with the measured spectra A by calculating the RMSE

In case a plurality of samples is available, a sample with the largest residue may be removed in order to obtain faster convergence. Optionally, also more samples may be removed.

· The measurement with the largest residue is removed from the set

• The Iteration is repeated (by removing the largest residue, the remaining set will converge to values resulting in smaller average residue)

The above algorithm indicates features and advantages that can be obtained using a method as described in an embodiment of the present invention.

In one aspect, the present invention also relates to a system for performing analysis of spectrophotometer data of one or more samples consisting of a number of constituents. The system comprises an input means for obtaining prior information for the one or more samples regarding their constituents and for obtaining UV-VIS spectrophotometer data for the one or more samples. The latter may be an input port for receiving prior information and spectrophotometer data. Alternatively the input means may also comprise a measurement system for recording UV-VIS spectrophotometer data such as for example a spectrophotometer. The system furthermore comprises a processing means programmed for defining a number of overlapping components contributing in the UV-VIS spectrophotometer data, the number of components comprising one or more components assigned to known constituents of the one or more samples and the number of components comprising one or more components that cannot be assigned to known constituents of the one or more samples, including one or more components representative for turbidity and/or scatter. Such a processing means may be for example a CPU although embodiments of the present invention are not limited thereto. The processing means according to embodiments of the present invention furthermore is programmed for using the prior information for the one or more samples regarding their constituents and using the UV-VIS spectrophotometer data, estimating the constituents composition and the component contributions to the UV-VIS spectrophotometer data for the number of components for the one or more samples by minimizing a residue between the UV-VIS spectrophotometer data and a fit based on said constituent composition and said component contributions thereby taking into account one or more components representative for turbidity and/or scatter, thus obtaining information regarding the one or more components that cannot be assigned to known constituents of the one or more samples and regarding the one or more components assigned to known constituents of the one or more samples. Such information may be putted out using an outputting means, such as e.g. a memory, a display, a printer or a plotter. Further features and advantages may correspond with components adapted for performing method steps as described in one or more of the methods according to embodiments of the aspects of the present invention.

In another aspect, the present invention relates to a method for extracting component contributions, e.g. reference spectra for particular constituents of the sample or for contributions of such constituents being modified, disturbed or being in a particular environment. The reference spectra can be extracted using methods as described above.

In one aspect, the present invention also relates to the use of a method as described above for particular applications. One of these applications may be identifying and/or quantifying contaminants in DNA or RNA samples. The contaminatns may be protein contamination. The contaminants in one application may be PCR-inhibiting contaminants. Another application envisaged is the quantification of an amount of double-stranded DNA in a mixture of double-stranded DNA and single-stranded RNA or DNA. Still a further application is determination of the composition of a polynucleotide or protein. A further application is the quantification of the modification efficiency in a sample. The method also may be applied for validating a sample composition or a constituent thereof or group of constituents thereof. By extension the method allows to extract more detailed information about the composition of a component, in particular of polymers, and more specifically in biopolymers such as polynucleotides and proteins. Inversely, the method also allows to extract quantitative information of polymers based on the a priori knowledge of the composition of the known components, rather than the knowledge of the spectrum of the component. An example thereof is the quantification of correctly synthesised oligonucleotides relying on the known composition of the oligo in a sample that is suspected to be contaminated either by incompletely synthesized oligo's or by oligonucleotides synthesized in a previous run.

In still another aspect, embodiments of the present invention also relate to computer-implemented methods for performing at least part of the methods for analyzing spectrophotometer data. Embodiments of the present invention also relate to corresponding computer program products. The methods may be implemented in a computing system. They may be implemented as software, as hardware or as a combination thereof. Such methods may be adapted for being performed on computer in an automated and/or automatic way. In case of implementation or partly implementation as software, such software may be adapted to run on suitable computer or computer platform, based on one or more processors. The software may be adapted for use with any suitable operating system such as for example a Windows operating system or Linux operating system. The computing means may comprise a processing means or processor for processing data. According to some embodiments, the processing means or processor may be adapted for analyzing spectra according to any of the methods as described above or extracting reference spectra according to any of the methods as described above. Besides a processor, the computing system furthermore may comprise a memory system including for example ROM or RAM, an output system such as for example a CD-rom or DVD drive or means for outputting information over a network. Conventional computer components such as for example a keybord, display, pointing device, input and output ports, etc also may be included. Data transport may be provided based on data busses. The memory of the computing system may comprise a set of instructions, which, when implemented on the computing system, result in implementation of part or all of the standard steps of the methods as set out above and optionally of the optional steps as set out above. The obtained results may be outputted through an output means such as for example a plotter, printer, display or as output data in electronic format.

Further aspect of embodiments of the present invention encompass computer program products embodied in a carrier medium carrying machine readable code for execution on a computing device, the computer program products as such as well as the data carrier such as dvd or cd-rom or memory device. Aspects of embodiments furthermore encompass the transmitting of a computer program product over a network, such as for example a local network or a wide area network, as well as the transmission signals corresponding therewith.

By way of illustration, embodiments of the present invention not being limited thereby, experimental results illustrating features and advantages of at least some embodiments according to the present invention are discussed below.

FIG. 4 illustrates the effect of turbidity on UV-VIS spectra recorded for a set of samples that are homogenized by shaking. As can be seen, the contribution of the turbidity to the spectrum results in a reduced a plurality of UV-VIS spectra recorded for a set of samples. The more pronounced the turbidity, the more difficult it is to distinguish the other contributions. For reasons of comparison, in FIG. 5 the UV-VIS absorption spectra for the same sample set is shown whereby the spectra are recorded after the samples have been centrifuged, resulting a reduced turbidity as can be seen by comparison.

FIG. 6 illustrates a deconvolution of UV-VIS spectral data recorded for a sample comprising double stranded dna. The deconvolution is based on a first fit in a region where substantially no other components are present, in the present example being in the wavelength range between 350nm and 550nm . The first fitting is performed using a linear combination of two correction components representative for turbidity. Each of the correction components is inversely proportional with an n-th power of a the wavelength. In the present example, the fitting is performed using a first and second correction component , whereby the first correction component is inversely proportional with the 1,8 ^th power of the wavelength, i.e. the first correction component oc

λ

and whereby the second correction component is inversely proportional with the 4 power of the wavelength, i.e. the second correction component

The deconvolution shown indicates the measured spectrum 602, the fitted spectrum 604, the residue 606, turbidity and/or scatter contribution 608, the double stranded dna spectrum 610, the RNA/nucleotides spectrum 612 and the EDTA contamination contribution 614.

FIG. 7 illustrates the deconvolution of the spectrum recorded for the same sample, but after centrifugation. The turbidity present in the sample is significantly lower, but still present, and is also compensated for using the above correction components for fitting. The different components contributing in both samples are double stranded DNA, RNA or separate nucleotides. Further a residue component and a turbidity component based on a linear combination of two correction components as described above are taken into account. The residue component obtained is in both cases below 3%, whereby as a measure the root mean square difference between the fitting and the measured spectrum is taken, relative to the integrated area below the measured spectrum.

It was found that for both deconvolutions performed, the amount of double stranded dna that could be derived was the same, within measurement error, and the amount of RNA/nucleotides component that could be derived was the same, within measurement error. The fact that the same results are obtained using the same sample under significantly different turbidity conditions, confirms that the correction performed for turbidity allow for obtaining accurate quantification using UV-VIS absorption spectroscopy.

Results on other experimental measurement data, result in the same advantages as described above.