Analysis method, evaluation method, and program

Abstract

The present disclosure discloses a method for analyzing a plurality of samples obtained in oligo-nucleic acid synthesis. The analysis method disclosed in the present disclosure includes implementing multivariate analysis using an analysis result pertaining to each of the plurality of samples. The analysis method furthermore includes outputting the analysis result pertaining to each of the plurality of samples in a mode conforming to a result of the multivariate analysis.

Description

Analysis method, evaluation method, and program

[0001] The present invention relates to an analytical method, an evaluation method, and a program, and in particular to a method for analyzing data of multiple samples obtained in oligonucleic acid synthesis, a method for evaluating the identity of impurity profiles, and a program for analyzing data of the multiple samples.

[0002] Various studies have been conducted on nucleic acid medicines. Japanese Patent No. 5569264 (Patent Document 1) discloses a technique using mass spectrometry for sequence analysis of RNA (ribonucleic acid) of 20 mer or more, which is the target of nucleic acid medicines.

[0003] Patent No. 5569264

[0004] Nucleic acid drugs are produced by chemical synthesis, but many impurities are generated during the process. To obtain the active ingredient of a nucleic acid drug, the compound produced during synthesis is generally purified, and this purification process requires compound analysis. The results of this analysis are used to monitor what impurities have been removed.

[0005] The results of the above analysis are also used to determine the range of cycles in continuous chromatography purification that will yield a purified product with a stable composition from among multiple cycles of purified products, and to determine the range of fractions in HPLC (High Performance Liquid Chromatography) that will yield a purified product with a stable composition from among multiple fractions. For these determinations, there is a need for information regarding the identity of the composition of the purified product.

[0006] The present invention was devised in light of the above-mentioned circumstances, and its purpose is to provide a technology for providing information regarding the identity of the composition of purified products in the analysis of samples containing nucleic acids for nucleic acid medicines.

[0007] An analytical method according to one aspect of the present disclosure is a method for analyzing multiple samples obtained in oligonucleic acid synthesis, and includes the steps of performing multivariate analysis using the analytical results of each of the multiple samples, and outputting the analytical results of each of the multiple samples in a manner according to the results of the multivariate analysis.

[0008] An evaluation method according to an aspect of the present disclosure is a method for evaluating a synthetic crude product obtained by synthesizing an oligonucleic acid, and includes a separation step of separating a sample containing the synthetic crude product, a division step of dividing the sample into multiple samples for analysis, an analysis step of analyzing each of the multiple samples by mass spectrometry using matrix-assisted laser desorption / ionization, an analysis step of performing multivariate analysis using the analytical data, and a presentation step of presenting an index for evaluating the identity of impurity profiles among the multiple samples.

[0009] According to one aspect of the present disclosure, a technique is provided for providing information regarding the identity of the composition of a purified product in the analysis of a sample containing nucleic acid for nucleic acid medicine.

[0010] 1 is a diagram showing the configuration of an analysis system in which the analysis method of the present disclosure is implemented. FIG. 2 is a diagram showing an example of data related to multiple fractions obtained in an example of batch purification. FIG. 3 is a diagram showing an example of the configuration of "batch collection fractions" in this embodiment. FIG. 4 is a diagram showing an example of the results of multivariate analysis of the analysis results of batches (1) to (8). FIG. 5 is a diagram showing another example of the results of multivariate analysis of the analysis results of batches (1) to (8). FIG. 6 shows the MALDI-MS analysis results of batches (1) to (6). FIG. 7 shows the MALDI-MS analysis results of batches (7) and (8). FIG. 8 is a diagram showing, of the multiple fractions obtained in a certain cycle in continuous purification, a group of fractions collected as purified products in that cycle and a group of fractions sent to the next cycle. FIG. 9 is a diagram showing an example of the results of multivariate analysis of the analysis results of fractions collected as purified products in cycles (1) to (10). FIG. 10 is a diagram showing an example of the results of multivariate analysis performed by further utilizing the analysis results of a reference standard of the target nucleic acid drug. 10 is a diagram showing an example of the results of multivariate analysis performed by further utilizing the analytical results of a sample considered to have an impurity profile similar to the impurity profile of some of the samples used to create graph 100. FIG. 11 is a diagram showing an example of the results of multivariate analysis performed by further utilizing the analytical results of a crude product obtained through continuous purification. FIG. 12 is a diagram showing an example of the results of multivariate analysis performed by further utilizing the analytical results of a standard product and a crude product obtained through continuous purification. FIG. 13 is a diagram showing an example of the results of multivariate analysis of the analytical results of the samples from cycles (1) to (9) and the crude product sample. FIG. 14 is a diagram showing an example of the discrimination results for the test subject in 8 using a discriminant. FIG. 15 is a diagram showing an example of the discrimination results for the sample from cycle (10) using a discriminant. FIG. 16 is a graph in which plots corresponding to the test subjects in FIGS. 15 and 16 are added to the graph of FIG. 14. FIG. 17 is a flowchart of an example of processing performed in a computer 1 for analysis using analytical results of multiple samples. FIG. 18 is a flowchart of another example of processing performed in a computer 1 for analysis using analytical results of multiple samples.

[0011] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In the drawings, the same or corresponding parts are designated by the same reference numerals, and description thereof will not be repeated.

[0012] [Analysis System] Figure 1 is a diagram showing the configuration of an analysis system in which the analysis method of the present disclosure is implemented. The analysis system includes an analysis device 20 that generates analysis results for multiple samples to be analyzed, and a computer 1 that analyzes the analysis results. In this embodiment, the computer 1 executes a given application program (hereinafter also referred to as an "analysis program") to perform multivariate analysis using the analysis results of multiple samples obtained in the synthesis of nucleic acid pharmaceuticals.

[0013] The "multiple samples" may be multiple fractions in the "batch purification" described below, or samples collected in each of multiple cycles of the "continuous purification" described below.

[0014] The hardware configuration of a computer 1 will be described with reference to Figure 1. The computer 1 includes a central processing unit (CPU) 10 that performs calculations and control, and storage devices that store programs and data. The storage devices include a hard disk drive (HDD) 11 and random access memory (RAM) 12. The storage devices may include a solid state drive (SSD) instead of or in addition to the HDD.

[0015] The HDD 11 non-temporarily stores programs executed by the CPU 10 and data (e.g., analysis results) used in the execution of the programs. The CPU 10 can also execute programs stored in storage devices other than the HDD 11. The RAM 12 temporarily stores data used during the execution of programs by the CPU 10. The RAM 12 functions as a temporary data memory used as a work area.

[0016] The computer 1 further includes a communication I / F 13 and an input / output interface (I / F) 14. The communication I / F 13 is, for example, a network card, and is an interface for the computer 1 to communicate with an external device. An example of the external device is an analysis device 20, which will be described later.

[0017] The input / output I / F 14 is an interface for input to or output from the computer 1. The input / output I / F 14 is connected to a display device 2 and an operation device 3. The display device 2 is, for example, a display. The operation device 3 accepts inputs including user instructions. The operation device 3 is, for example, configured by a keyboard and / or a mouse.

[0018] In one implementation example, the analysis results of the above-mentioned multiple samples are transmitted from the analysis device 20 to the computer 1. The analysis results do not need to be transmitted directly from the analysis device 20 to the computer 1. For example, the analysis results may be input by a user to the computer 1 via the operation device 3.

[0019] "Analysis results of multiple samples" means results obtained by a common analytical method for each of the "multiple samples."

[0020] The analytical device 20 may be a device that performs any type of analytical technique. One example of the analytical device 20 is a high performance liquid chromatograph (HPLC) for high performance liquid chromatography. Another example of the analytical device 20 is a photodiode array detector (PDA) used in high performance liquid chromatography. In this case, the analytical results are, for example, the retention time, peak intensity, and / or peak width of a peak in a chromatogram.

[0021] The analysis results may be information about only the peaks of the impurities of interest among the peaks detected in the chromatogram (an example of data corresponding to impurities). "Impurities" are not limited to impurities derived from the synthesis pathway of nucleic acid drugs, but may also be completely different types of impurities that have been mixed in from elsewhere.

[0022] In this embodiment, it is not essential that the impurities be identified, that is, the sample to be analyzed by the analysis device 20 may contain unidentified compounds (impurities).

[0023] Yet another example of the analytical device 20 is a liquid chromatography mass spectrometer (LCMS) for liquid chromatography mass spectrometry, in which the analytical results are, for example, the m / z values ​​of peaks and / or peak intensities in a mass spectrum.

[0024] Yet another example of the analytical device 20 is a MALDI-MS (Matrix-Assisted Laser Desorption / Ionization-Mass Spectrometer) for mass spectrometry using matrix-assisted laser desorption / ionization. In this case, the analytical result may be the mass spectrum itself, a list of m / z values ​​of peaks extracted from the mass spectrum, or a set of m / z values ​​of peaks and their peak intensities (and / or peak areas). The analytical result may also be data obtained by performing a given processing such as deconvolution on the mass spectrum. The analytical result may also be a peak list of only the impurities of interest in the mass spectrum.

[0025] Another example of the analysis device 20 is an electrophoresis device. In this case, the analysis result is, for example, electrophoretic mobility. The analysis method performed in the electrophoresis device may be any method, such as capillary electrophoresis or microchip electrophoresis. The measurement time in an electrophoresis device is shorter than that in other types of devices such as HPLC. Therefore, in the analysis of this embodiment, using the analysis results of the electrophoresis device as the analysis result is preferable in terms of shortening the time required to prepare data for analysis.

[0026] The computer 1 may use a combination of results output from multiple analytical devices as the analysis result of the sample for analysis.

[0027] [Data Analysis of Multiple Samples in Batch Purification] In the synthesis of nucleic acid pharmaceuticals, one example of a method for purifying the obtained crude product is batch purification, in which the crude product is purified in a single separation using a single column.

[0028] In the context of batch purification, the term "sample" refers to a single fraction or a mixture of multiple fractions obtained in the separation using a single column. The term "sample" in the context of batch purification will be explained below.

[0029] 2 is a diagram showing an example of data relating to a plurality of fractions obtained in an example of batch purification. In this specification, as a specific example of purification, an example is adopted in which a nucleic acid drug having the same nucleic acid sequence as nusinersen is synthesized as an oligonucleic acid, and the crude synthesis is purified using an anion exchange column.

[0030] In Figure 2, line L10 represents the absorbance of ultraviolet light at a given wavelength for each of the multiple fractions. In the graph of Figure 2, the horizontal axis represents retention time, and the vertical axis represents absorbance. Line L10 also represents the absorbance of ultraviolet light at a given wavelength at that retention time. In the example of Figure 2, the start times of collection of the 12th to 32nd fractions are indicated by "F12" to "F32." Fraction F12 refers to the fraction collected from the retention time indicated by "F12" to the retention time indicated by "F13." Fraction F13 refers to the fraction collected from the retention time indicated by "F13" to the retention time indicated by "F14."

[0031] FIG. 3 shows an example of the structure of a "batch collection fraction" in this embodiment. In one example of data analysis, one "batch collection fraction" is defined by a single fraction or a mixture of multiple fractions. In this specification, the "batch collection fraction" may be described as "batch (1)" or the like.

[0032] In the example of Figure 3, batch (1) is defined by the mixture of fractions F13 and F14, and batches (2) to (8) are defined by the mixtures of fractions F15 to F17, F18 to F19, F20 to F24, F25 to F27, F28 to F31, F18 to F27, and F19 to F27, respectively.

[0033] 3, in batch purification, the retention times of the samples constituting each of the multiple batches are different from one another. In this sense, the purification aspects of the multiple batches are different from one another.

[0034] Figure 4 shows an example of the results of multivariate analysis of the analysis results of batches (1) to (8). Graph 40 in Figure 4 is a score plot showing the results of principal component analysis. In graph 40, the horizontal axis represents the first principal component, and the vertical axis represents the second principal component.

[0035] In Fig. 4, "batch" is shown as "Bt." In Fig. 4, the analysis results of each of the multiple samples in batch (1) are shown as multiple plots located near the string "Bt(1)."

[0036] The results in Figure 4 further employ a coarse body as the "sample." In Figure 4, the coarse body is shown as "Crude(1)." The plot of the coarse body is the plot located near the string "Crude(1)."

[0037] In graph 40, the plots of batches (4), (7), and (8) can be distinguished from the plots of other samples (batches (1) to (3), (5), and (6), and the coarse sample). Note that the plots of batches (4), (7), and (8) overlap with the plots of batch (3) and batch (5).

[0038] Therefore, by referring to graph 40, the user can determine that the impurity profiles of the samples from batches (3) to (5), (7), and (8) are consistent, and can therefore decide to collect the samples from batches (3) to (5), (7), and (8) as the purified products of the batch purification.

[0039] Figure 5 shows another example of the results of multivariate analysis of the analysis results of batches (1) to (8). Graph 50 in Figure 5 is a score plot showing the results of partial least squares discriminant analysis. The results in Figure 5, like those in Figure 4, use a more coarse sample as the "sample."

[0040] In graph 50, similar to graph 40, the plots of batches (4), (7), and (8) can be distinguished from the plots of other samples (batches (1) to (3), (5), and (6) and the coarse sample). Note that the plots of batches (4), (7), and (8) overlap with the plots of batch (3) and batch (5).

[0041] Therefore, by referring to graph 50, the user can determine that the impurity profiles of the samples from batches (3) to (5), (7), and (8) are consistent, and can therefore decide to collect the samples from batches (3) to (5), (7), and (8) as the purified products of the batch purification.

[0042] The determination of the identity of the impurity profiles based on Graph 40 or Graph 50 is also supported by the analytical results of each sample. Figure 6 shows the MALDI-MS analytical results for batches (1) to (6). Figure 7 shows the MALDI-MS analytical results for batches (7) and (8). In Figures 6 and 7, the horizontal axis represents m / z, and the vertical axis represents the detected intensity.

[0043] As shown in FIGS. 6 and 7, in the mass spectra of all batches (1) to (8), relatively large peaks appear near m / z=3562 and m / z=7125.

[0044] On the other hand, in the m / z range other than near m / z = 3562 and m / z = 7125, batches (3) to (5), (7), and (8) differ significantly from batches (1), (2), and (6). For example, in the m / z range of 3564 to 7125, only relatively small peaks appear in batches (3) to (5), (7), and (8), while at least one relatively large peak appears in batches (1), (2), and (6) and the crude product. From this, it is believed that the identity of the impurity profiles between samples is ensured among batches (3) to (5), (7), and (8), but the identity of the impurity profiles between batches (3) to (5), (7), and (8) and batches (1), (2), and (6) is not believed to be ensured.

[0045] Therefore, the judgment regarding the identity of the impurity profiles based on graph 40 or graph 50 is supported by the MALDI-MS analysis results shown in FIGS.

[0046] 4 and 5, the range of the plots may be set as a condition for the identity of the impurity profiles by statistical processing of the plots in the score plots. In one implementation example, information specifying the condition is stored in the HDD 11. One example of the set condition may be that the center of gravity of all plots in the score plot is specified, and a figure of a given size (for example, an ellipse having a major axis in a first direction and a minor axis in a second direction) centered on the center of gravity contains a number of plots equal to or greater than a given percentage.

[0047] In the analysis system, the computer 1 may select a plot that satisfies the above conditions from the score plots, and select a set of batches that have identity based on the selection.

[0048] [Data Analysis of Multiple Samples in Continuous Purification] Another example of a method for purifying a crude product obtained in the synthesis of nucleic acid pharmaceuticals is continuous purification. In continuous purification, the crude product is purified by continuous chromatographic separation in multiple stages. In continuous purification, separation at a certain stage is defined as one cycle, and the crude product is finally purified after multiple cycles of separation. In one implementation example, a twin column such as the "Contichrom CUBE" manufactured by YMC Co., Ltd. (https: / / www.ymc.co.jp / chromato / contichrom / ) is used for continuous purification. In the following description, the first to tenth cycles in an example of continuous purification are referred to as cycles (1) to (10), respectively. In continuous purification, cycles (1) to (10) differ from each other in the number of separations on the column. In this sense, the purification modes of cycles (1) to (10) are different from each other.

[0049] FIG. 8 shows a group of fractions collected as purified products in a cycle of continuous purification and a group of fractions sent to the next cycle, among the multiple fractions obtained in that cycle.

[0050] In the example of Figure 8, fractions F18 to F27 are collected as purified products, as shown as "Collection." Furthermore, fractions F13 to F17 and F28 to F30 are sent to the next cycle, as shown as "Recycle." Note that in the next cycle, a necessary amount of crude material is sent in addition to fractions F13 to F17 and F28 to F30, so that the same amount of separation target sent to the column in the previous cycle is sent to the column in the next cycle.

[0051] In continuous purification, the fractions collected as purified product in each cycle (fractions F18 to F27) are an example of a “sample.” “Samples” from multiple cycles are an example of “multiple samples.”

[0052] Figure 9 shows an example of the results of multivariate analysis of the analysis results of fractions collected as purified products in cycles (1) to (10). Graph 90 in Figure 9 is a score plot showing the results of partial least squares discriminant analysis. In Figure 9, "cycle" is shown as "Cycle." In Figure 9, the analysis results of each of the multiple samples in cycle (1) are shown as multiple plots located near the string "Cycle(1)."

[0053] In graph 90, the plots for cycles (2) to (10) and the plots for cycle (1) can be distinguished.

[0054] Therefore, by referring to graph 90, the user can determine that the impurity profiles of the samples from cycles (2) to (10) are consistent, and can therefore decide to use the samples from cycles (2) to (10) as the final purified product of the continuous purification process.

[0055] 9 , a plot range may be set as a condition for the identity of the impurity profile by statistical processing of the plots in the score plot. In one implementation example, information specifying the condition is stored in the HDD 11. One example of the set condition may be that the center of gravity of all plots in the score plot is specified, and a figure of a given size (for example, an ellipse having a major axis in a first direction and a minor axis in a second direction) centered on the center of gravity contains a number of plots equal to or greater than a given percentage.

[0056] In the multivariate analysis using the analytical results of the samples from cycles (1) to (10), analytical results of other compounds may also be used.

[0057] Figure 10 shows an example of the results of multivariate analysis performed by further utilizing the analytical results of a standard preparation of the nucleic acid drug of interest. Graph 100 shown in Figure 10 shows the results of multivariate analysis performed by utilizing the analytical results of the fractions collected as purified products in cycles (1) to (10) as well as the analytical results of the standard preparation of the nucleic acid drug of interest. As an example of multivariate analysis, partial least squares discriminant analysis is employed. In graph 100, the plot of the standard preparation is shown as multiple plots near the string "Std."

[0058] The standard may be, for example, a highly purified commercial product, and is more pure than the purified product collected in the continuous production. Therefore, it is expected that the impurity profile in the samples from cycles (2) to (10) will be different from that of the standard.

[0059] 10, in addition to graph 100, graph 90 is also shown for reference. In graph 100, the plots for cycles (2) to (100) are located relatively close to each other, but are separated from the plot for the standard product. This indicates that the difference in impurity profile between cycles (2) to (10) and the standard product is reflected in the results of multivariate analysis.

[0060] In graph 100, the range in which the plots for cycles (2) to (10) are present is narrower than in graph 90. This allows the multivariate analysis to more reliably distinguish cycles (2) to (10) from other samples by utilizing the analytical results of a standard sample whose impurity profile differs from that of cycles (2) to (10).

[0061] Figure 11 shows an example of the results of multivariate analysis performed by further utilizing the analytical results of a sample believed to have an impurity profile similar to the impurity profile of some of the samples used to create graph 100. Graph 110 shown on the right side of Figure 11 is the result of multivariate analysis performed by utilizing the analytical results of batch (7) described with reference to Figure 3, in addition to the analytical results of a standard and fractions collected as purified products in cycles (1) to (10). In addition to graph 110, graph 100 is also shown on the left side of Figure 11 for reference. Batch (7) represents a sample collected as purified product in batch purification.

[0062] In graph 110, like graph 100, the plots for cycles (2) to (100) are positioned in a way that allows them to be distinguished from the remaining plots.

[0063] Batch (7) represents a sample collected as a purified product as described above. Cycle (1) represents a sample collected in the first cycle. Therefore, it is assumed that the impurity profiles of the samples in batch (7) and cycle (1) are identical or similar. In graph 110, the plots for cycle (1) and batch (7) are located close to each other. This supports the assumption that multivariate analysis of analytical results reflects the identity (approximation) of impurity profiles among multiple samples.

[0064] Figure 12 shows an example of the results of multivariate analysis performed using the analytical results of the crude product from the continuous purification. Graph 120 shown in Figure 12 shows the results of multivariate analysis performed using the analytical results of the crude product in addition to the analytical results of the fractions collected as purified products in cycles (1) to (10). In graph 120, the plot of the crude product is shown as multiple plots near the character string "Crude."

[0065] 12 also shows the results of a multivariate analysis performed using the analytical results of batch (7) described above as graph 121. The multivariate analysis shown in graph 121 utilizes the analytical results of batch (7) described above in addition to the analytical results used in the multivariate analysis of graph 120.

[0066] In the graph 120, the plots of cycles (2) to (10) are located at positions that can be distinguished from the plot of cycle (1), and also at positions that can be distinguished from the plot of the coarse body.

[0067] In graph 121, similar to graph 120, the plots of cycles (2) to (10) are located in positions that are distinct from the plot of cycle (1) and also distinct from the plot of the coarse product. Furthermore, in graph 121, the plot of cycle (1) and the plot of the coarse product are located close to each other. This supports the assumption that the multivariate analysis of analytical results reflects the similarity (approximation) of impurity profiles among multiple samples.

[0068] 13 is a diagram showing an example of the results of multivariate analysis performed using the analytical results of the standard and the crude product of the continuous purification. Graph 130 shown in FIG. 13 shows the results of multivariate analysis performed using the analytical results of the standard and the crude product in addition to the analytical results of the fractions collected as purified products in cycles (1) to (10).

[0069] In graph 130, the plots for cycles (2) to (10) are in positions that can be distinguished from the plots for cycle (1), the standard, and the crude material.

[0070] In graph 130, the distance between the range where the plots for cycles (2) to (10) are present and the range where the plot for cycle (1) is present, and the distance between the range where the plots for cycles (2) to (10) are present and the range where the plot for the standard product is present, are both shorter than the distance between the range where the plots for cycles (2) to (10) are present and the range where the plot for the crude product is present. On the other hand, the difference between the impurity profile of the samples for cycles (2) to (10) and the impurity profile of the sample for cycle (1), and the difference between the impurity profile of the samples for cycles (2) to (10) and the impurity profile of the standard product sample, are both assumed to be smaller than the difference between the impurity profile of the samples for cycles (2) to (10) and the impurity profile of the crude product. These facts support the assumption that multivariate analysis of analytical results reflects the identity (approximation) of impurity profiles between multiple samples.

[0071] [Generation of Collection Conditions by Machine Learning] The multivariate analysis method used in the analysis system is not limited to statistical methods. A machine learning method may be used as the multivariate analysis method. Furthermore, the analysis system may generate conditions for samples to be collected as purified products. The generation of such conditions will be described with reference to FIGS. 14 to 17 .

[0072] In this example, samples from cycles (1) to (10) are collected as purified products. Figure 14 shows an example of the results of multivariate analysis of the analysis results of samples from cycles (1) to (9) and the crude sample. Principal component analysis is used as an example of multivariate analysis.

[0073] In graph 140 of FIG. 14 , the plots of samples (1) to (9) are shown as filled circles near the character string "purified." The plots of the coarse samples are shown as filled triangles near the character string "coarse." As can be seen from graph 140, the plots of "purified" are located in an area separated from the area where the plots of "coarse." This allows samples (1) to (9) to be distinguished from the coarse samples by multivariate analysis.

[0074] In this example, training data is generated by tagging the analysis results of the samples from cycles (1) to (9) as "purified" and the analysis results of the crude sample as "crude." Then, a machine learning process using the training data generates a discriminant that determines whether input data is "purified" or "crude." The machine learning process uses, for example, a random forest algorithm. The analysis results may be, for example, a list of peaks in a mass spectrum (the m / z values ​​and intensities of one or more peaks).

[0075] Fig. 15 is a diagram showing an example of discrimination results for eight test subjects using the discriminant equation. The eight test subjects to which the discrimination results shown in Fig. 15 correspond were obtained from coarse samples different from the coarse samples used to generate the discriminant equation. The discrimination results in Fig. 15 were obtained by applying the analysis results of each of the eight test subjects to the discriminant equation. Fig. 15 shows the discrimination results as "Group," and further shows the score of each discrimination result as "Score."

[0076] As shown in Fig. 15, the discrimination results for all eight test objects were "coarse body." Since eight test objects should be discriminated as "coarse body," all of the discrimination results for the eight test objects shown in Fig. 15 are correct.

[0077] Fig. 16 is a diagram showing an example of the discrimination results for the samples in cycle (10) using the discriminant equation. The eight test subjects to which the discrimination results shown in Fig. 16 correspond were obtained from the samples in cycle (10). The discrimination results in Fig. 15 were obtained by applying the analysis results of each of the eight test subjects to the discriminant equation. Fig. 16 shows the discrimination results as "Group," and further shows the score of each discrimination result as "Score."

[0078] As shown in Figure 16, the discrimination results for all eight test subjects were "purified." Since the eight test subjects should have been discriminated as "purified," the discrimination results for the eight test subjects shown in Figure 16 were all correct.

[0079] From the above, a discriminant generated by machine learning processing using the analysis results of multiple samples can be used to classify unknown samples as "purified" or "crude." Samples classified as "purified" are samples that should be collected as purified products.

[0080] Therefore, the above-mentioned discriminant equation can be used as a condition (condition of identity of impurity profile) for a sample to be collected as a purified product relative to an unknown sample.

[0081] Figure 17 is a graph of Figure 14 with plots added corresponding to the test subjects of Figures 15 and 16. In graph 170 of Figure 17, the plots 8 corresponding to the test subjects of Figure 15 are shown as open triangle plots on the "crude" side, and the plots 8 corresponding to the test subjects of Figure 16 are shown as open circle plots on the "purified" side.

[0082] In graph 170, the open triangle plots are located near the filled triangle plots. Also, the open square plots are located near the filled square plots. As a result, the results of the principal component analysis can be used as conditions for distinguishing between "purified" and "crude" unknown samples, similar to the discriminant equations described above.

[0083] [Processing Flow (1)] FIG. 18 is a flowchart of an example of processing performed in the computer 1 for analysis using the analysis results of a plurality of samples.

[0084] In one implementation example, the process is performed by the CPU 10 of the computer 1 executing a given program (analysis program). For example, while the analysis program is running, the user may input, as an analysis instruction, the type of sample to be used in the analysis (cycles (1) to (10) only, cycles (1) to (10) and a crude sample, cycles (1) to (10) and a crude sample and a standard, etc.). The process of FIG. 18 may be started in response to the user inputting information instructing the analysis program to start the analysis.

[0085] In step S10, the computer 1 reads out an analysis instruction, and in step S20, the computer 1 reads out from the HDD 11 the analysis results of the sample type specified in the analysis instruction.

[0086] In step S30, the computer 1 performs multivariate analysis using the analysis results read out in step S20.

[0087] In step S40, the computer 1 outputs the results of the multivariate analysis performed in step S30. The output may be a display of the results on the display device 2 or a transmission of the results to an external device. The output results represent the analysis results in a format according to the results of the multivariate analysis. The output results may be, for example, a score plot as shown in FIG. 4 or the like.

[0088] Thereafter, the computer 1 ends the processing of Fig. 18. [Processing Flow (2)] Fig. 19 is a flowchart of another example of processing for analysis using the analysis results of multiple samples, which is performed by the computer 1. Compared to the processing of Fig. 18, the processing of Fig. 19 further includes control of steps S32 to S36.

[0089] More specifically, after the computer 1 performs the multivariate analysis in step S30, the control proceeds to step S32.

[0090] In step S32, the computer 1 generates the conditions for the identity of the impurity profiles.

[0091] In step S34, the computer 1 extracts samples that satisfy the condition generated in step S32 from the samples that were the subject of the multivariate analysis in step S30. For example, if the "condition" generated in step S32 specifies a given range on the score plot, in step S34, samples having plots that fall within that range are extracted.

[0092] In step S36, the computer 1 stores the conditions generated in step S32 in the HDD 11. Thereafter, the computer 1 advances control to step S40.

[0093] The results output in the process of FIG. 19 may include the conditions generated in step S32 and / or the samples extracted in step S34. The conditions may be output as a graphic representing a range on the score plot shown in FIG. 4. The extracted samples may be output in a different display format from the non-extracted samples on the score plot. For example, the plots of the extracted samples may be displayed with a higher density or as larger graphics than the plots of the non-extracted samples.

[0094] Thereafter, the computer 1 ends the processing of Fig. 19. As described above, the results of the multivariate analysis reflect differences in the impurity profiles. Based on this, in this embodiment, conditions for the identity of the impurity profiles among multiple samples are generated based on the results of the multivariate analysis of the multiple samples.

[0095] The conditions stored in step S36 may be used in subsequent processing. For example, the conditions generated during the production of a drug for clinical trials may be used in production after the clinical trials to ensure that the impurity profile of the produced drug is identical to the impurity profile of the drug for clinical trials. The conditions stored in step S36 may also be used to confirm the identity of the currently purified lot and the previously purified lot.

[0096] [Index for Evaluating Identity of Impurity Profiles] The analysis system according to the present disclosure may present the above-described "conditions" as indices for evaluating identity of impurity profiles.

[0097] In this example, the analysis system may employ a MALDI-MS as the analysis device 20, and fractions from a liquid chromatograph may be introduced into the MALDI-MS. The liquid chromatograph separates a plurality of fractions from a sample containing a synthetic crude. The analysis device 20 derives analysis results for each of the plurality of fractions. The computer 1 then performs multivariate analysis using the analysis results to generate the above-mentioned "conditions" as indicators for evaluating the identity of the impurity profiles. The impurities contained in at least one of the plurality of fractions include, for example, at least one of desulfurized products, deletion products, adducts, and products in which purine bases have been eliminated from the oligonucleic acid that is the target of synthesis.

[0098] The computer 1 may display the above index on a display device (for example, the display device 2) as a graphic representation of the index in a score plot such as that described with reference to FIG. 4 and the like.

[0099] The computer 1 may use the index to determine whether the impurity profiles of multiple samples satisfy identity and output the determination result. For example, when a certain index (condition) is identified from samples of a certain group, the computer 1 may determine whether samples outside the group satisfy identity with respect to samples constituting the group as follows. That is, the computer 1 may define a shape (e.g., an ellipse) identified from the index (condition) in the score plot, and determine that identity is satisfied when the plot of the sample outside the group is located within the shape. On the other hand, the computer 1 may determine that identity is not satisfied when the plot of the sample outside the group is not located within the shape.

[0100] The index (condition) may be generated using analytical results of fractions collected from a standard.

[0101] In this example, the synthetic crude product may be purified by batch purification or continuous purification, i.e., the fraction whose analytical results were used to generate the above-mentioned indicators (conditions) may be sent to the next cycle in continuous purification.

[0102] Aspects It will be understood by those skilled in the art that the exemplary embodiments described above are examples of the following aspects.

[0103] (Item 1) An analytical method according to one aspect is a method for analyzing a plurality of samples obtained in oligonucleic acid synthesis, and may include the steps of: performing multivariate analysis using the analytical results of each of the plurality of samples; and outputting the analytical results of each of the plurality of samples in a manner according to the results of the multivariate analysis.

[0104] According to the analytical method described in paragraph 1, a technique is provided for providing information regarding the identity of the composition of a purified product in the analysis of a sample containing nucleic acid for nucleic acid medicine.

[0105] (Item 2) In the analysis method described in item 1, each of the plurality of samples may contain an impurity.

[0106] According to the analytical method described in the second paragraph, knowledge regarding the identity of impurity profiles in a plurality of samples can be obtained by multivariate analysis.

[0107] (Item 3) In the analysis method according to item 1 or 2, the analysis results may include data corresponding to impurities.

[0108] According to the analytical method described in Section 3, knowledge regarding the identity of impurity profiles in a plurality of samples can be obtained by multivariate analysis.

[0109] (Item 4) In the analysis method according to item 3, the impurities may include unidentified compounds.

[0110] The analytical method described in Section 4 makes it possible to provide information regarding the identity of the composition among multiple samples without the need to identify impurities.

[0111] (Item 5) In the analysis method according to any one of items 1 to 4, the analysis results may include results of at least one of high performance liquid chromatography, liquid chromatography mass spectrometry, mass spectrometry using matrix-assisted laser desorption / ionization, and electrophoresis.

[0112] According to the analysis method described in Section 5, the results of an existing analysis method can be used as the analysis results.

[0113] (Item 6) In the analysis method according to any one of items 1 to 5, the aspect according to the result of the multivariate analysis may include displaying by score plot.

[0114] According to the analysis method described in item 6, the identity of the composition of the purified product can be output in an easily recognizable manner.

[0115] (Item 7) The analysis method according to any one of items 1 to 6 may further include a step of generating conditions for the identity of impurity profiles based on the results of the multivariate analysis.

[0116] The analytical method described in item 7 provides an objective condition for the identity of impurity profiles based on the results of multivariate analysis of multiple samples.

[0117] (Item 8) The analysis method according to any one of items 1 to 7 may further include the step of extracting a sample that satisfies the condition from the plurality of samples.

[0118] According to the analytical method described in item 8, one or more samples having the same impurity profile are extracted from a plurality of samples.

[0119] (Item 9) In the analysis method according to any one of Items 1 to 8, the plurality of samples may include at least one of a standard sample of the oligonucleic acid and a crude product obtained by the synthesis of the oligonucleic acid.

[0120] According to the analytical method described in item 9, the analytical results of at least one of the standard and crude sample are utilized in the multivariate analysis.

[0121] (Item 10) In the analysis method according to any one of Items 1 to 9, the plurality of samples may include two or more samples obtained by the oligonucleic acid synthesis and purified in different ways.

[0122] According to the analytical method described in item 10, information regarding the identity of impurity profiles in two or more samples that have been purified in different ways is provided.

[0123] (Item 11) An evaluation method according to one aspect is a method for evaluating a synthetic crude product obtained by synthesizing an oligonucleic acid, and may include a separation step of separating a sample containing the synthetic crude product, a division step of dividing the sample into a plurality of samples for analysis, an analysis step of analyzing each of the plurality of samples by mass spectrometry using matrix-assisted laser desorption / ionization, an analysis step of performing multivariate analysis using analytical data, and a presentation step of presenting an index for evaluating the identity of impurity profiles among the plurality of samples.

[0124] According to the evaluation method described in item 11, a technique is provided for providing information on the identity of the composition of a purified product in the analysis of a sample containing nucleic acid for use in nucleic acid medicines.

[0125] (12) In the evaluation method according to the 11th aspect, each of the plurality of samples may contain an impurity.

[0126] According to the evaluation method described in item 12, information can be provided that takes into consideration the identity of impurity profiles in multiple samples.

[0127] (Item 13) In the evaluation method according to Item 12, the impurities may include at least one of desulfurized products, deletion products, adducts, and products in which purine bases have been eliminated from the oligonucleic acid.

[0128] According to the evaluation method described in paragraph 13, when the impurities include at least one of desulfurized products, deletion products, adducts, and products in which purine bases have been removed from the oligonucleic acid, information can be provided that takes into account the identity of the impurity profiles in multiple samples.

[0129] (14) In the evaluation method according to any one of paragraphs 11 to 13, the presenting step may include displaying the index in a score plot.

[0130] According to the evaluation method described in paragraph 14, the index can be displayed in a form that is easy to recognize intuitively.

[0131] (Item 15) The evaluation method according to any one of items 11 to 14 may further include a determination step of determining whether or not the impurity profiles among the plurality of samples satisfy identity based on the index.

[0132] According to the evaluation method described in item 15, more information based on the index can be provided. (Item 16) In the evaluation method described in any one of items 11 to 15, the plurality of samples may include a standard preparation of the oligonucleic acid.

[0133] According to the evaluation method described in paragraph 16, the index can be generated taking into account the characteristics of the standard. (Item 17) The evaluation method described in any one of paragraphs 11 to 15 may further comprise a purification step of purifying the sample thus collected.

[0134] The evaluation method described in paragraph 17 can provide an indication of samples that can be further purified.

[0135] (Item 18) A program according to one aspect may be executed by a computer to cause the computer to carry out the analysis method according to any one of items 1 to 10.

[0136] According to the program described in item 18, a technique is provided for providing information regarding the identity of the composition of a purified product in the analysis of a sample containing nucleic acid for use in nucleic acid medicine.

[0137] The embodiments disclosed herein should be considered to be illustrative in all respects and not restrictive. The scope of the present disclosure is defined by the claims, not by the description of the above-described embodiments, and is intended to include all modifications within the meaning and scope of the claims. Furthermore, it is intended that each technique in the embodiments can be implemented alone or, if necessary, in combination with other techniques in the embodiments to the extent possible.

[0138] 1 Computer, 2 Display device, 3 Operation device, 10 CPU, 20 Analysis device

Claims

1. A method for analyzing a plurality of samples obtained in the synthesis of oligonucleic acids, comprising: analyzing the plurality of samples by liquid chromatography mass spectrometry (LCMS); acquiring a plurality of analytical results corresponding to each of the plurality of samples from the LCMS; the plurality of analytical results include chromatograms and spectra; performing a multivariate analysis using the plurality of analysis results; and outputting the plurality of analysis results in a manner according to the result of the multivariate analysis.

2. The multivariate analysis is a principal component analysis for obtaining a first principal component and a second principal component, the first principal component is retention time; The analysis method according to claim 1 , wherein the second principal component is a spectrum.

3. The analytical method according to claim 1 , wherein each of the plurality of samples includes an impurity.

4. 3. The analysis method according to claim 1, wherein the analysis results include data corresponding to impurities.

5. The analytical method according to claim 4 , wherein the impurities include unidentified compounds.

6. The analysis method according to claim 1 or 2, wherein the manner of performing the multivariate analysis includes displaying the results by a score plot.

7. The analysis method according to claim 1 or 2, further comprising the step of generating conditions for the identity of impurity profiles based on the results of the multivariate analysis.

8. The analysis method according to claim 7 , further comprising the step of extracting a sample that satisfies the condition from the plurality of samples.

9. 3. The analysis method according to claim 1, wherein the plurality of samples include at least one of a standard sample of the oligonucleic acid and a crude product obtained by synthesis of the oligonucleic acid.

10. 3. The analysis method according to claim 1, wherein the plurality of samples include two or more samples obtained by synthesis of the oligonucleic acid and purified in different ways.

11. A method for evaluating a crude oligonucleotide obtained by synthesis, comprising: a fractionation step of fractionating a sample containing the synthetic crude material; dividing the sample into multiple samples for analysis; an analyzing step of analyzing each of the plurality of samples by mass spectrometry using matrix-assisted laser desorption / ionization; an analysis step of performing multivariate analysis using the analysis data; and a presentation step of presenting an index for evaluating the identity of impurity profiles among the plurality of samples.

12. The evaluation method according to claim 11 , wherein each of the plurality of samples includes an impurity.

13. The evaluation method according to claim 12, wherein the impurities include at least one of desulfurized products, deletion products, adducts, and products in which a purine base has been eliminated from the oligonucleic acid.

14. The evaluation method according to claim 11 or 12, wherein the presenting step includes displaying the index in a score plot.

15. 13. The evaluation method according to claim 11, further comprising a determining step of determining whether or not the impurity profiles among the plurality of samples satisfy identity based on the index.

16. The evaluation method according to claim 11 or 12, wherein the plurality of samples include a standard sample of the oligonucleic acid.

17. The evaluation method according to claim 11 or 12, further comprising a purification step of purifying the sample.

18. A program that, when executed by a computer, causes the computer to carry out the analysis method according to claim 1 or 2.