Information processing device, information processing method, and computer program

Abstract

To provide an information processing device, an information processing method, and a computer program capable of analyzing data related to genetic information.SOLUTION: An information processing device of this embodiment includes a processing unit that analyzes data related to genetic information in accordance with the purpose of the data analysis to obtain analysis result data, generates a prompt for input to a language model that instructs the analysis result data to be interpreted based on gene-related information related to the analysis result data and the purpose of the data analysis, and obtains an interpretation result of the analysis result data based on the prompt and the language model.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present embodiment relates to an information processing device, an information processing method, and a computer program. [Background technology]

[0002] Integrated analysis of multiple datasets related to genetic information is a useful technique for disease analysis, biomarker discovery, and other purposes. This type of integrated analysis is called meta-analysis. Because genetic information is measured values ​​derived from living organisms, the number of samples in a single dataset is relatively small, and integrated analysis of multiple datasets is often required. In the field of genetic information, information processing devices and methods specialized for meta-analysis have been developed.

[0003] However, in order to use information processing devices and methods specialized for meta-analysis, advanced expertise in genetic information has been required to search and select datasets and apply analytical techniques. Furthermore, biological and statistical knowledge has been required to interpret the results of meta-analysis. Thus, conducting a meta-analysis requires a lot of time and effort, making it difficult for non-experts to perform. Therefore, there is a need for a method that allows meta-analysis to be performed in a short period of time without requiring advanced expertise in genetic information. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2006-294014 [Patent Document 2] Patent No. 5822309 [Patent Document 3] Patent No. 5261644 Summary of the Invention [Problem to be solved by the invention]

[0005] The present embodiment provides an information processing device, an information processing method, and a computer program capable of analyzing data related to genetic information. [Means for solving the problem]

[0006] The information processing device of this embodiment includes a processing unit that analyzes data related to genetic information in accordance with the purpose of the data analysis to obtain analysis result data, generates a prompt for inputting a language model that instructs the analysis result data to be interpreted based on gene-related information related to the analysis result data and the purpose of the data analysis, and obtains an interpretation result of the analysis result data based on the prompt and the language model. [Brief explanation of the drawings]

[0007] [Figure 1] FIG. 1 is a block diagram showing an example of the configuration of an information processing apparatus according to an embodiment of the present invention. [Figure 2] FIG. 2 is a block diagram showing an example of the configuration of a dataset search unit according to the embodiment of the present invention. [Figure 3] 10A to 10C are diagrams showing examples of selected datasets and their evaluation results, and selection information of the selected datasets according to an embodiment of the present invention. [Figure 4] FIG. 2 is a block diagram showing an example of the configuration of a meta-analysis unit according to an embodiment of the present invention. [Figure 5] 10A and 10B are diagrams showing examples of analysis results and processed analysis results according to an embodiment of the present invention. [Figure 6] FIG. 2 is a block diagram showing an example of the configuration of an interpretation generating unit according to the embodiment of the present invention. [Figure 7] FIG. 10 is a diagram illustrating a template of a prompt for generating an interpretation according to an embodiment of the present invention. [Figure 8] 10A and 10B are diagrams illustrating examples of prompts for generating an interpretation according to an embodiment of the present invention. [Figure 9] FIG. 10 is a diagram showing an example of interpretation information according to the embodiment of the present invention. [Figure 10] FIG. 10 is a diagram showing an example of user presentation data presented to a user. [Figure 11] 1 is a flowchart showing an outline of a procedure in an information processing device according to an embodiment of the present invention. [Figure 12] 10 is a flowchart showing an example of a procedure for generating interpretation information for analysis results according to an embodiment of the present invention. [Figure 13] FIG. 1 is a hardware block diagram of an information processing apparatus according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0008] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0009] 1 is a block diagram showing a schematic configuration of an information processing device 100 according to an embodiment of the present invention. The information processing device 100 according to this embodiment includes a processing unit 1, a data input unit 10 (input unit), and a data output unit 50 (output unit). The processing unit 1 includes a dataset search unit 20, a meta-analysis unit 30, and an interpretation generation unit 40. The dataset search unit 20 includes a public dataset DB 205, which is a third database (see FIG. 2), as a database (DB), and the interpretation generation unit 40 includes a gene-related information DB 404, which is a first database, and a gene-related additional information DB 406, which is a second database (see FIG. 6).

[0010] The data input unit 10 accepts various instructions or information from the user as input. For example, it accepts dataset search instruction information as instructions or information for the dataset search unit 20. The search instruction information includes both or at least one of information on the purpose of the analysis and search condition information. The data input unit 10 also accepts information specifying a dataset to be selected from multiple datasets acquired by the dataset search unit 20 (dataset selection information), and information specifying the addition of a dataset not acquired by the dataset search unit 20 (dataset addition instruction information). The data input unit 10 may also accept information indicating an instruction to generate an interpretation to the interpretation generation unit 40 (interpretation generation instruction information) from the user.

[0011] The method and format by which the data input unit 10 receives this information may be a text input in which the user inputs characters using a keyboard, an input in which the user uses a mouse to click on options or buttons on the screen, an input in which the user uses a touch panel to tap on options or buttons on the screen, an input in which the user uses a microphone to input characters by voice, an input in which the user uses a camera to input characters by moving their face or hands, an input in which the user uses a pen to input characters by hand, or the like.

[0012] As an example for explaining this embodiment, the information on the purpose of analysis is a sentence such as "I would like to search for microRNA biomarkers that are effective in distinguishing pancreatic cancer." The search condition information in this example includes information indicating the condition that the GEO (Gene Expression Omnibus) database be used and information indicating the condition that the measurement device be limited to a microarray.

[0013] The dataset search unit 20 generates a search statement based on the search instruction information acquired from the data input unit 10, and uses the search statement to search the public dataset DB 205 to acquire multiple datasets. For example, if the search condition information includes a condition for using a specific database, the search unit searches for databases that satisfy the condition. If the search condition information includes a condition for specifying a measuring device, the search targets datasets for which measurements were performed using the measuring device. The dataset search unit 20 evaluates the acquired multiple datasets, selects two or more databases, and outputs the selected datasets (selected datasets). Note that in this embodiment, two or more databases are selected, but it is not excluded to select only one database depending on the purpose of the data analysis.

[0014] FIG. 2 is a block diagram showing an example of the configuration of the dataset search unit 20. The dataset search unit 20 includes a query generation unit 201, a dataset search unit 202, a dataset evaluation unit 203, a dataset selection unit 204, and a public dataset DB 205.

[0015] The search query generation unit 201 generates a search query based on the search instruction information. More specifically, the search query is generated based on at least one of the analysis objective information and the search condition information included in the search instruction information.

[0016] The search statement is a keyword or query used to search the public dataset DB 205. Here, the text of the acquired information to be analyzed may be used as the search statement as is, or the text of the acquired information to be analyzed may be added to the text of the search condition information to use as the search statement. The search statement may be generated by processing information using a rule-based method, or by processing information using a statistical method or a machine learning method.

[0017] In the example mentioned above, the keywords "pancreatic cancer" and "microRNA," which are related to genetic information, are extracted from the sentence "I would like to search for microRNA biomarkers that are effective in identifying pancreatic cancer." The search query "pancreatic cancer AND microRNA AND microarray" is generated so that it includes the extracted keywords and information on the measurement device and matches the search query in the GEO database.

[0018] The public dataset DB 205 stores or accumulates publicly available datasets (public datasets). For example, the public datasets include the identification ID, title, summary, number of specimens, measurer, measurement target, measurement device, measurement values, related research, etc. of the public dataset.

[0019] The dataset search unit 202 instructs a search in the public dataset DB 205 using the search statement generated by the search statement generation unit 201, and acquires multiple searched datasets. In the example described above, the dataset search unit 202 uses the search statement "pancreatic cancer AND microRNA AND microarray" to search the GEO database, which is the public dataset DB 205, and acquires multiple datasets as search results.

[0020] The dataset evaluation unit 203 evaluates multiple datasets acquired by the search. The evaluation of the datasets may be performed based on attribute information of the datasets or based on the measurement data itself. The evaluation criteria may be the quality and reliability of the meta-analysis, the availability of the data, or the relevance to the information on the purpose of the analysis or the search condition information. The evaluation criteria may be changed depending on the purpose of the meta-analysis and the search condition information.

[0021] In the example described above, the dataset evaluation unit 203 evaluates multiple datasets by extracting the genes to be measured by the measuring instruments included in the datasets and the number of specimens measured, and calculates the number of overlapping genes to be measured between the multiple datasets. It also checks whether the same dataset (overlapping dataset) exists between the multiple datasets. It may also check the number of specimens measured in each dataset.

[0022] The dataset selection unit 204 determines selection criteria according to the above evaluation criteria, and selects those datasets that satisfy the selection criteria from the multiple datasets evaluated by the dataset evaluation unit 203. Information on the selected dataset (selected dataset) is output to the user via the data output unit 50. The dataset information output to the user includes, for example, the dataset name and information extracted or calculated for evaluation by the dataset evaluation unit 203. The dataset information output to the user may or may not include the measurement data itself.

[0023] The dataset selection unit 204 may acquire, from the data input unit 10, at least one of information indicating multiple datasets selected by the user from the selected datasets (dataset selection information) and information on datasets to be added (dataset addition information). When the dataset selection information is acquired, the set of selected datasets is updated by selecting two or more datasets from the selected datasets (by removing datasets that were not selected). When the dataset addition information is acquired, the set of selected datasets is updated by adding the datasets indicated in the dataset addition information to the selected datasets. The datasets indicated in the dataset addition information may be provided directly by the user or may be acquired from the public dataset DB 205. When the dataset selection unit 204 receives neither dataset selection information nor dataset addition information from the user, it adopts the selected datasets selected by itself.

[0024] The dataset selection unit 204 may adopt the selected dataset as is without outputting it to the user. In this way, the process related to the selection of the dataset may be performed solely by the dataset selection unit 204 based on an algorithm provided in the dataset selection unit 204, or part or all of the process may be performed based on the user's judgment.

[0025] In the example described above, for example, the dataset selection unit 204 eliminates overlapping datasets and selects datasets with a threshold or more of measured specimens (deletes datasets with a number of measured specimens less than the threshold). Also, it selects multiple datasets (two or more) so that the number of overlapping genes to be measured is a threshold or more.

[0026] The dataset selection unit 204 automatically removes duplicate datasets and selects datasets with a number of measured samples equal to or greater than a threshold (deletes datasets with a number of measured samples less than the threshold), and outputs the resulting datasets to the user. The user may select two or more datasets from these output datasets so that the number of duplicated genes to be measured does not fall below the threshold, and input the dataset selection information indicating the selected datasets. In this way, some or all of the processes related to dataset selection may be performed based on user instructions.

[0027] Figure 3(A) is a table showing information about selected datasets after duplicate datasets have been eliminated and datasets with a measurement sample count below a threshold have been deleted, along with an example of evaluation information for each dataset. The figure shows the dataset name, sample count, measurement target genes, and a summary. The sample count and the number of duplicates are examples of evaluation information for the datasets. The process of selecting two or more datasets so that the number of duplicates of measurement target genes is equal to or greater than a threshold has not yet been performed.

[0028] FIG. 3(B) is an example of dataset selection information entered by a user who has viewed the information in FIG. 3(A) to narrow down the datasets to be used in the analysis. The dataset selection information indicates that datasets A and B are to be used, and datasets C and D are not to be used. Specifically, the user excludes dataset C from use because it contains liver cancer, which is not a target. Furthermore, excluding dataset D, the number of overlapping genes to be measured between dataset A and dataset B is three (gene 1, gene 2, gene 3). For this reason, the user has decided not to use datasets C and D in the analysis. Note that this decision may be made by the dataset selection unit 204 rather than by the user.

[0029] FIG. 4 is a block diagram showing an example of the configuration of the meta-analysis unit 30. As shown in FIG. The meta-analysis unit 30 processes the data for analysis from the selected dataset obtained from the dataset search unit 20, applies meta-analysis techniques using the processed data, processes the meta-analysis results, which are the analysis result data, for use as input to a language model, and outputs the processed results.

[0030] The meta-analysis unit 30 includes an analytical data processing unit 301, a meta-analysis application unit 302, and a language model input data processing unit 303.

[0031] The analytical data processing unit 301 extracts necessary data or information from the selected dataset, processes the data for applying meta-analysis, and generates processed data. Specifically, it performs data integration, data shaping, and processing of missing values ​​and outliers. The extracted data may be gene statistics or effect sizes in addition to gene measurement values.

[0032] In this example, the expression levels of microRNA measured by microarray are extracted, and numerical processing (logarithmic processing) is performed to match the scale between datasets. Specifically, for the two datasets, dataset A and dataset B shown in Figure 3(B), the M included in dataset A is A The measured values ​​of gene expression levels of gene 1, gene 2, and gene 3 of the samples are extracted, and the M B The measured values ​​of the gene expression levels of gene 1, gene 2, and gene 3 for each sample are extracted, and the data in dataset B are logarithmically processed to match the scale.

[0033] The meta-analysis application unit 302 acquires the processed data from the analytical data processing unit 301, applies a meta-analysis method, and generates analysis results. Meta-analysis methods include integration methods such as fixed-effect models and random-effect models, and visualization methods such as histograms and forest plots. Any one of these methods or other methods may be used, or multiple methods may be combined. The meta-analysis application unit 302 performs data analysis by executing analysis algorithms that realize these methods.

[0034] In this example, for each gene, the hypothesis is whether the expression level differs between pancreatic cancer specimens and non-cancer specimens. Statistical values ​​indicating whether the gene is differentially expressed in each dataset are calculated, and a P value is calculated by integrating these values ​​and performing a null hypothesis test. Specifically, for each gene g = 1 to 3, the statistical values ​​indicating whether the gene is differentially expressed between pancreatic cancer specimens and non-cancer specimens are calculated as δ g,A In dataset B, δ g,B The P value P of each gene g is calculated by integrating these and testing the null hypothesis. g Calculate.

[0035] Figure 5(A) is a table showing an example of the analysis results. A table containing the P-value for each gene is shown.

[0036] The language model input data processing unit 303 acquires the meta-analysis results from the meta-analysis application unit 302, processes them for language model input, and generates processed analysis results. Specifically, it extracts key information from tables, figures, etc. showing the analysis results, and converts them into a format that can be input to a language model such as natural language. Generally, when processing data for language model input, information processing may be performed using a rule-based method, or information processing may be performed using a statistical method or a machine learning method.

[0037] In this example, a table containing a list of differentially expressed microRNAs (gene 1, gene 3) and their P-values ​​(P1, P3) between pancreatic cancer and non-cancer specimens is generated in text format. Figure 5(B) is a table showing an example of processed analysis results. The table includes the P-values ​​of gene 1 and gene 3. In this example, gene 2 has a P-value P2 that is less than the significance level α, so it is not estimated to be differentially expressed at or above the significance level α and is therefore not included in the table.

[0038] The interpretation generation unit 40 generates an interpretation generation prompt including gene-related information related to the processed analysis results based on the processed analysis results obtained from the meta-analysis unit 30 and information on the purpose of the analysis. At this time, if interpretation generation instruction information (described later) from the user is obtained from the data input unit 10, the interpretation generation prompt may also be used to generate the interpretation generation prompt. The interpretation generation unit 40 uses the generated interpretation generation prompt and a language model to generate and output interpretation information corresponding to the meta-analysis results.

[0039] FIG. 6 is a block diagram showing an example of the configuration of the interpretation generating unit 40. As shown in FIG. The interpretation generation unit 40 includes an interpretation generation prompt generation unit 401, a language model interpretation generation unit 402, a data processing unit for user presentation 403, a gene-related information DB 404 (first database), a gene-related additional information search unit 405, and a gene-related additional information DB 406 (second database).

[0040] The gene-related information DB 404 stores or accumulates information about genes, including, for example, gene sequence information and functions, gene transcription products and their functions, and relationships between genes. Related information about each gene can be obtained from the gene-related information DB 404.

[0041] The interpretation generation prompt generation unit 401 acquires the processed analysis results from the meta-analysis unit 30, and generates an interpretation generation prompt including gene-related information related to the processed analysis results based on the processed analysis results and information on the purpose of the analysis. The information on the purpose of the analysis may be acquired via the meta-analysis unit 30, or may be extracted from search instruction information input from the data input unit 10.

[0042] The prompt generation unit 401 selects an appropriate template (prompt template for interpretation generation) depending on the information to be analyzed and the processed analysis results, etc. The template has a format adapted to the language model used in the language model interpretation generation unit 402.

[0043] Figure 7 shows an example of a prompt template for generating an interpretation. The example in Figure 7 is a prompt template for generating an interpretation corresponding to the results of a meta-analysis of microRNAs, with the variables being "processed analysis results," "gene-related information," "additional gene-related information," and "information on the purpose of the analysis." As an example, "processed analysis results" corresponds to the first variable, "gene-related information" or "additional gene-related information" corresponds to the second variable, and "information on the purpose of the analysis" corresponds to the third variable. "Information on the purpose of the analysis" is included in the response instructions for the language model. Text other than the variables is already included in the template. If the user provides instructions for generating an interpretation, a different template may be used.

[0044] The interpretation generation prompt generating unit 401 acquires gene-related information related to the processed analysis results from the gene-related information DB 404 .

[0045] The gene-related additional information search unit 405 searches the gene-related additional information DB 406 based on the processed analysis results from the meta-analysis unit 30 to obtain other types of gene-related information related to the processed analysis results. The gene-related additional information DB 406 stores different types of gene-related information from the gene-related information DB 404. The gene-related additional information DB 406 is, for example, a DB that is updated daily. The gene-related information obtained from the gene-related additional information DB 406 is particularly referred to as gene-related additional information. The gene-related additional information search unit 405 outputs the obtained gene-related additional information to the interpretation generation prompt generation unit 401.

[0046] The gene-related information DB 404 and the gene-related additional information DB 406 are separate DBs, but may be the same DB. According to this embodiment, the database that stores information related to genes includes at least one of the gene-related information DB 404 and the gene-related additional information DB 406.

[0047] The interpretation generation prompt generation unit 401 replaces the "gene-related information" and "gene-related additional information" variables in the template with the gene-related information acquired from the gene-related information DB 404 and the gene-related additional information acquired from the gene-related additional information DB 406. In this way, an intermediate interpretation generation prompt is generated in which some of the variables in the interpretation generation template have been replaced.

[0048] The language model interpretation generation unit 402 acquires an intermediate interpretation generation prompt from the interpretation generation prompt generation unit 401, and replaces the "processed analysis result" and "information on the purpose of analysis" among the variables in the acquired interpretation generation prompt with the processed analysis result and information on the purpose of analysis, thereby generating an interpretation generation prompt.

[0049] FIG. 8 shows an example of a prompt for generating an interpretation generated by the prompt generation unit 401 for generating an interpretation. In the example of FIG. 8, background knowledge about microRNAs is included in the text as shown (note that in this example, the background knowledge is included in the template in advance). The content of FIG. 5(B) is stored as the analysis result. As gene-related information, three names and representative functions of biomolecules that are presumed targets of Gene 1 and Gene 3 are included. In addition, as other types of gene-related information, two titles and summaries of papers related to Gene 1 and Gene 3 are included. In addition, instructions for response to the language model are included in the text as shown.

[0050] The language model interpretation generation unit 402 inputs the generated interpretation generation prompt into a language model to generate interpretation information. The interpretation information is interpretation information corresponding to the information targeted for analysis and the processed analysis results, and is result information of the answer requested by the answer instruction included in the interpretation generation prompt. The language model is a processing device that can understand the context and meaning from natural language sentences and generate new sentences using natural language processing technology. The language model may be a large-scale language model. However, the input and output of the language model are not limited to natural language sentences and may also include images, etc.

[0051] In this embodiment, the execution engine of the language model is provided in the language model interpretation generation unit 402, but the execution engine of the language model may exist as an external server, and interpretation information may be received from the server by sending an interpretation generation prompt to the server. The language model may be a general-purpose language model or a language model adjusted for interpretation generation purposes.

[0052] Fig. 9 shows an example of information for interpretation of the analysis result generated by the language model interpretation generation unit 402. In the example of Fig. 9, the information for interpretation of the analysis result includes the content and consideration of the analysis result.

[0053] The data processing unit 403 for user presentation evaluates the quality of the interpretation information and determines whether to have the prompt generating unit 401 regenerate the prompt for interpretation generation.

[0054] The quality is evaluated, for example, as follows. · It is determined whether the number of characters in the interpretation information is equal to or greater than a threshold value. If it is less than the threshold value, it may be determined that the amount of information is small and a decision may be made to regenerate the information. A language model may be used to generate a numerical value for evaluating the quality of the interpretation information, and the evaluation may be performed based on whether the numerical value is equal to or greater than a threshold value. Alternatively, the quality may be evaluated using a rule-based method, a statistical method, or a machine learning method.

[0055] When it is decided to regenerate the interpretation generation prompt, the data processing unit 403 for user presentation sends instruction information (interpretation generation instruction information) for regenerating the interpretation generation prompt to the interpretation generation prompt generating unit 401 to reacquire interpretation information. The interpretation generation prompt used at this time may be the same as the previous one (even an interpretation generation prompt with the same content may result in a different answer (interpretation information)). Alternatively, the gene-related information DB 404 or the gene-related additional information DB 406 may be searched again to acquire gene-related information or gene-related additional information (for example, by increasing the number of items acquired in the search from the previous time), and the reacquired gene-related information or gene-related additional information may be additionally included in the previous interpretation generation prompt (it is not necessary to add the same information as the previous one). Note that if the gene-related information DB 404 or the gene-related additional information DB 406 is updated over time, gene-related information or gene-related additional information that was not found in the previous search may be acquired in the current search. Alternatively, new information (additional information) may be added to the previous interpretation generation prompt. The additional information may be generated in accordance with the case where the user inputs interpretation generation instruction information, which will be described later. The generation of interpretation information is repeated until the quality satisfies a termination condition (for example, exceeds a threshold value or reaches a maximum number of repetitions).

[0056] The data-for-user presentation processing unit 403 processes the processed analysis results and interpretation information into a format that can be presented to the user, corresponding to the information of the purpose of the analysis, to obtain data-for-user presentation. For example, data including the processed analysis results of FIG. 5(B) and the interpretation information of FIG. 9 is generated as data-for-user presentation. The data-for-user presentation may be presented to the user via the data output unit 50. The data-for-user presentation presented to the user may also include other information, such as information of the purpose of the analysis.

[0057] 10 shows an example of the data presented to the user, including information on the purpose of the analysis, processed graph data of the analysis results, and interpretation information.

[0058] If the user checks the data to be presented to the user and is not satisfied with the interpretation information, the user may input interpretation generation instruction information from the data input unit 10 and cause the language model interpretation generation unit 402 to generate interpretation information again. In this case, the interpretation generation prompt generation unit 401 generates an interpretation generation prompt again based on the user's interpretation generation instruction information and sends the interpretation generation prompt to the language model interpretation generation unit 402, causing the language model interpretation generation unit 402 to generate interpretation information again.

[0059] The user's interpretation generation instruction information may, for example, instruct the user to include the interpretation generation results of a previous query with the same content in the most recently generated prompt for interpretation generation. This may allow the user to comment on the differences between the current consideration (interpretation result) and the previous consideration (interpretation result) in the regenerated interpretation information. If the user has additional knowledge or questions about the considerations included in the interpretation information, the user may regenerate the interpretation information by adding the additional knowledge or questions to the prompt for interpretation generation. This also enables the generation of an interpretation in an interactive format. The user may also use the interpretation generation instruction information to re-search the gene-related information DB 404 or the gene-related additional information DB 406 to obtain gene-related information or gene-related additional information (e.g., to increase the number of results obtained in the search compared to the previous time), and to additionally include the re-obtained gene-related information or gene-related additional information in the previous prompt for interpretation generation (it is not necessary to add the same information as the previous time). If the gene-related information DB 404 or the gene-related additional information DB 406 is updated over time, the current search may obtain gene-related information or gene-related additional information that was not found in the previous search.

[0060] The data output unit 50 outputs information on the selected dataset (see Figure 3(A)), dataset evaluation results (see "Number of overlaps" in Figure 3(A) and "Do not use" or "Use" in Figure 3(B)), processed analysis results (Figure 5(B)), interpretation information (Figure 9), data to be presented to the user, etc.

[0061] The data output unit 50 may have any configuration for outputting this information or data as long as it is capable of allowing the user to recognize the information or data. For example, the data output unit 50 may be a display such as a liquid crystal display device or an organic EL display device, and may display the information or data using visual elements such as images or natural language. The data output unit 50 may be a printer, and may print the information or data using visual elements such as images or natural language. The data output unit 50 may be an application, and may save the information or data in a file in an appropriate file format. Furthermore, the data output unit 50 may be a communication device capable of sending and receiving information or data to and from a user's terminal, a server, or the like, and may transmit the information or data via a network.

[0062] 11 is a flowchart outlining the procedure in the information processing device according to the first embodiment of the present invention. The following is just one example, and the dataset search, meta-analysis, and generation of interpretation information do not necessarily have to be performed exactly as shown.

[0063] In step S101, the input unit 10 receives search instruction information as input from the user. The search instruction information includes, for example, information on the purpose of analysis and search condition information.

[0064] In step S102, the dataset search unit 20 generates a search statement using at least one of the analysis target information and the search condition information, and uses the generated search statement to search the public dataset DB 205. The dataset evaluation unit 203 evaluates the retrieved multiple datasets and selects multiple datasets based on the evaluation results.

[0065] In step S103, the output unit 50 outputs information about the selected data set and information about the evaluation result to the user (see FIG. 3(A)).

[0066] In step S104, the input unit 10 receives from the user information (database selection information) instructing the selection (narrowing down) of the selected datasets or information instructing the addition of a database (database addition instruction information). The database addition instruction information is instruction information input by the user when he / she wishes to further narrow down the datasets from the output selected datasets. The database addition instruction information is instruction information input by the user when he / she wishes to add an additional dataset. The dataset selection unit 204 updates the set of selected datasets based on one or both of these pieces of instruction information. When database selection information is input, the number of selected datasets decreases, and when database addition instruction information is input, the number of selected datasets increases. Note that the user may not input either piece of instruction information.

[0067] In step S105, the meta-analysis unit 30 processes the selected dataset for analysis, applies a meta-analysis method to the processed data to obtain meta-analysis results, and processes the meta-analysis results for language model input. For example, necessary information is extracted from the selected dataset information, and the data is processed for applying meta-analysis to generate processed data. For example, the expression levels of microRNAs are extracted and numerical processing (logarithmic processing) is performed to match the scale between datasets. For each gene, statistics indicating whether the gene is differentially expressed in each dataset are calculated based on the hypothesis that the expression level differs between pancreatic cancer specimens and non-cancer specimens, and a P-value is calculated by integrating these statistics and performing a null hypothesis test (see FIG. 5(A)). A table containing a list of genes differentially expressed between pancreatic cancer specimens and non-cancer specimens and the P-values ​​for each gene is generated in text format (see FIG. 5(B)).

[0068] In step S106, the interpretation generation unit 40 generates an interpretation generation prompt that includes gene-related information and the like related to the processed analysis results. If interpretation generation instruction information is input, the information specified by the interpretation generation instruction information is also included in the interpretation generation prompt (see FIG. 8). Using the interpretation generation prompt and the language model, interpretation information (see FIG. 9) corresponding to the information targeted by the analysis and the processed analysis results is generated.

[0069] In step S107, the output unit 50 outputs the processed analysis results (see FIG. 5(B)) and interpretation information (see FIG. 9) to the user.

[0070] In step S108, the process returns to step S106 in which the data input unit 10 has received interpretation generation instruction information from the user as input, and if not, the process ends.

[0071] Fig. 12 is a flowchart showing an example of detailed operations in step S106 in Fig. 11. This flowchart shows in more detail an example of the procedure for generating information for interpreting the analysis results.

[0072] In step S1061, the gene-related additional information search unit 405 retrieves other types of gene-related information (additional gene-related information) related to the processed analysis results from the gene-related additional information DB 406. For example, it searches for papers related to the genes (gene 1 and gene 3) included in the processed analysis results and retrieves two topics and summaries for each.

[0073] In step S1062, if there is interpretation generation instruction information from the user or interpretation generation instruction information from the data processing unit for user presentation 403, the prompt generation unit for interpretation generation 401 acquires the interpretation generation instruction information from the data input unit 10 or the data processing unit for user presentation 403.

[0074] In step S1063, the interpretation generation prompt generating unit 401 selects an appropriate template according to at least one of the information to be analyzed and the processed analysis results.

[0075] In step S1064, the interpretation generation prompt generation unit 401 acquires gene-related information related to the processed analysis results from the gene-related information DB 404. The acquired gene-related information and the additional gene-related information acquired in step S1061 are added to the template. Specifically, the acquired gene-related information and the additional gene-related information acquired in step S1061 replace corresponding variables in the selected template. If interpretation generation instruction information has been acquired, content specified by the interpretation generation instruction information may be further added to the template. Adding content specified by the interpretation generation instruction information also makes it possible to generate an interpretation in an interactive format.

[0076] In step S1065, the analysis objective information and the processed analysis results are added to the appropriate locations in the template. Specifically, the analysis objective information and the processed analysis results are substituted for the corresponding variables in the template. Steps S1064 and S1065 replace all variables in the template (adding information necessary for interpretation generation to the template), thereby obtaining a prompt for interpretation generation. Note that part of the process for generating the prompt for interpretation generation may be performed by the language model interpretation generation unit 402. For example, the process of adding the analysis objective information and the processed analysis results to the template may be performed by the language model interpretation generation unit 402.

[0077] In step S1066, the user-presented data processing unit 403 acquires interpretation information from the language model interpretation generation unit 402 and evaluates the quality of the interpretation information based on the information to be analyzed and the processed analysis results. For example, the evaluation may be based on the number of characters in the interpretation information, or a numerical value for evaluating the quality of the interpretation information may be generated using a language model. Alternatively, the quality may be evaluated using a rule-based method, a statistical method, or a machine learning method.

[0078] In step S1067, if the quality exceeds a predetermined threshold or reaches a predetermined maximum number of repetitions, proceed to step S1068. Otherwise, user presentation data processing unit 403 generates interpretation generation instruction information that instructs regeneration of an interpretation generation prompt, and returns to step S1062.

[0079] In step S1068, the data processing unit for user presentation 403 processes the processed analysis results and interpretation information into a format that can be presented to the user, corresponding to the information targeted for analysis, and presents the processed results to the user via the data output unit 50.

[0080] As described above, according to this embodiment, meta-analysis can be automated without requiring advanced specialized knowledge of genetic information, and analysis can be performed in a short period of time.

[0081] (Variation 1) The information to be analyzed may be "microRNA associated with colorectal cancer in Japanese people," "DNA sequence associated with type 2 diabetes in Japanese people," "RNA sequence associated with cerebral ischemic disease," "mRNA sequence associated with psychiatric disorders," etc. The search condition information may be information indicating the condition to use the "ENA database" or "DDBJ database" as the search database, or information indicating the condition to use "PCR" or "NGS" as the measurement device.

[0082] (Variation 2) The language model interpretation generation unit 402 may generate data processing instruction information in addition to the interpretation information, and the user-presentation data processing unit 403 may process data such as the processed analysis results in accordance with the data processing instruction information to generate additional interpretation information and output the additional interpretation information together with the interpretation information. For example, the language model interpretation generation unit 402 may generate, in addition to a document representing the interpretation information as interpretation information, instruction information (data processing instruction information) for processing data such as the processed analysis results to generate a graph showing that information. The user-presentation data processing unit 403 may generate a graph in accordance with the data processing instruction information and output user-presentation data that combines the document representing the interpretation information with the graph.

[0083] (Variation 3) In the template of Figure 7, the knowledge item already contains information ("1. MicroRNAs are non-coding RNAs with very short base lengths, and have been reported to show specific expression patterns in certain cancer cells."), but knowledge may also be treated as a variable, allowing information to be embedded in the template. In this case, for example, the gene-related information DB404 may be accessed from the information contained in the query entered by the user, and related information corresponding to that information may be obtained and embedded in the knowledge item of the template. For example, if the query contains "microRNA," the gene-related information DB404 may be accessed for microRNA, and the corresponding related information may be obtained and embedded in the knowledge item of the template.

[0084] (Hardware configuration) 13 shows the hardware configuration of the information processing device 100. The information processing device 100 is configured by a computer device 600. The computer device 600 includes a CPU 601, an input interface 602, a display device 603, a communication device 604, a main memory device 605, and an external memory device 606, which are interconnected by a bus 607.

[0085] The CPU (Central Processing Unit) 601 executes an information processing program, which is a computer program, on the main memory device 605. The information processing program is a program that realizes each of the above-mentioned functional components of the information processing device 100. The information processing program may be realized not by a single program, but by a combination of multiple programs and scripts. Each functional component is realized by the CPU 601 executing the information processing program.

[0086] The input interface 602 is a circuit for inputting operation signals from input devices such as a keyboard, a mouse, and a touch panel to the information processing device 100. The input interface 602 corresponds to the data input unit .

[0087] The display device 603 displays data output from the information processing device 100. The display device 603 is, for example, but not limited to, an LCD (liquid crystal display), an organic electroluminescence display, a CRT (cathode ray tube), or a PDP (plasma display). Data output from the computer device 600 can be displayed on the display device 603. The display device 603 corresponds to the data output unit 50.

[0088] The communication device 604 is a circuit that enables the information processing device 100 to communicate with an external device wirelessly or via a wire. Data can be input from the external device via the communication device 604. The data input from the external device can be stored in the main memory device 605 or the external memory device 606. The communication device 604 corresponds to the data output unit 50.

[0089] The main memory device 605 stores an information processing program, data required for executing the information processing program, data generated by executing the information processing program, etc. The information processing program is deployed and executed on the main memory device 605. The main memory device 605 is, for example, a RAM, a DRAM, or an SRAM, but is not limited to these. Each storage unit or database of the information processing device 100 may be constructed on the main memory device 605.

[0090] The external storage device 606 stores information processing programs, data required for executing the information processing programs, data generated by executing the information processing programs, etc. These information processing programs and data are read into the main storage device 605 when the information processing programs are executed. The external storage device 606 is, for example, but is not limited to, a hard disk, an optical disk, a flash memory, or a magnetic tape. Each storage unit or database of the information processing device 100 may be constructed on the external storage device 606.

[0091] The information processing program may be pre-installed in the computer device 600, or may be stored in a storage medium such as a CD-ROM. The information processing program may also be uploaded onto the Internet.

[0092] Furthermore, the information processing device 100 may be configured as a single computer device 600, or may be configured as a system made up of a plurality of computer devices 600 connected to each other.

[0093] The present invention is not limited to the above-described embodiments, and the components can be modified and embodied in practice without departing from the spirit of the invention. Furthermore, various inventions can be created by appropriately combining multiple components disclosed in the above-described embodiments. For example, configurations in which some components are omitted from all the components shown in each embodiment may also be considered. Furthermore, components described in different embodiments may be appropriately combined.

[0094] This embodiment can also have the following configuration. [Item 1] Analyze the genetic information data according to the purpose of the data analysis to obtain analysis result data; a processing unit that generates a prompt for inputting a language model that instructs interpretation of the analysis result data based on gene-related information associated with the analysis result data and a purpose of the data analysis, and obtains an interpretation result of the analysis result data based on the prompt and the language model; An information processing device comprising: [Item 2] the processing unit acquires a template for creating the prompt, the template including a first variable for storing the analysis result data, a second variable for storing the gene-related information related to the analysis result data, and a third variable for storing a purpose of the data analysis, and generates the prompt by storing the analysis result data in the first variable of the template, the gene information in the second variable, and the purpose of the data analysis in the third variable. Item 1. An information processing device according to item 1. [Item 3] the processing unit evaluates the quality of the interpretation result, regenerates the prompt in accordance with the evaluation result, and obtains an interpretation result of the analysis result data based on the regenerated prompt and the language model. Item 3. The information processing device according to item 1 or 2. [Item 4] an output unit that outputs the interpretation result; an input unit for inputting information, the processing unit acquires additional information for the prompt from the input unit, and regenerates the prompt further based on the acquired additional information. Item 3. An information processing device according to item 3. [Item 5] The additional information obtained includes a user question regarding the interpretation result; the processing unit obtains an answer to the question together with an interpretation result of the analysis result data based on the regenerated prompt and the language model. Item 5. An information processing device according to item 4. [Item 6] The acquired additional information includes interpretation results previously acquired for data analysis with the same purpose as the purpose of the data analysis; the processing unit acquires information describing a difference between the interpretation result of the analysis result data and the previously acquired interpretation result, based on the regenerated prompt and the language model. Item 6. The information processing device according to item 4 or 5. [Item 7] the processing unit searches a first database storing a plurality of pieces of gene-related information based on the analysis result data, thereby acquiring the gene-related information related to the analysis result data; 7. The information processing device according to any one of items 1 to 6. [Item 8] the processing unit searches a second database that stores a plurality of pieces of other gene-related information different from the gene-related information based on the analysis result data, thereby acquiring the other gene-related information related to the analysis result data; generating the prompt further based on the acquired additional gene-related information; Item 8. An information processing device according to item 7. [Item 9] The analysis result data includes a plurality of genes according to the purpose of the data analysis. 9. The information processing device according to any one of items 1 to 8. [Item 10] The gene-related information includes information on genes that are putative targets of the genes included in the analysis result data. Item 10. The information processing device according to item 9. [Item 11] the processing unit generates the prompt based on other gene-related information related to the analysis result data that is different from the gene-related information; The other gene-related information is information on a paper related to the gene included in the analysis result data. Item 11. An information processing device according to item 10. [Item 12] The processing unit performs the data analysis by executing an analysis algorithm according to the purpose of the data analysis. 12. The information processing device according to any one of items 1 to 11. [Item 13] The data analysis is a meta-analysis. Item 13. An information processing device according to item 12. [Item 14] The data related to the genetic information is a dataset including measurement data related to genes for a plurality of specimens. 14. The processing device according to any one of items 1 to 13. [Item 15] the processing unit acquires the data related to the genetic information by searching a third database that stores a plurality of data related to the genetic information, based on at least one of a purpose of the data analysis and a search condition for the data related to the genetic information. 15. The information processing device according to any one of items 1 to 14. [Item 16] the search conditions include at least one of a condition specifying a database to be used for the search and a condition specifying a measuring device used to measure the genetic information; the processing unit searches a database that satisfies the search conditions, or searches for data in which the genetic information is measured by the measurement device. Item 16. An information processing device according to item 15. [Item 17] the data related to the genetic information and each of the data in the third database are datasets including measurement data related to genes for a plurality of specimens, the processing unit acquires a plurality of data sets from the third database; selecting a plurality of datasets from the plurality of acquired datasets based on the number of specimens and the genes to be measured in each of the plurality of acquired datasets, and analyzing the selected plurality of datasets to obtain the analysis result data; Item 17. The information processing device according to item 15 or 16. [Item 18] an output unit that outputs information on the selected plurality of data sets; an input unit for inputting information, the processing unit acquires, from the input unit, selection information for selecting a part of the selected plurality of data sets or addition instruction information for instructing addition of a new data set; updating a set of the selected plurality of datasets based on the selection information or the additional instruction information, and analyzing the datasets of the updated set to obtain the analysis result data; Item 18. An information processing device according to item 17. [Item 19] Analyze the genetic information data according to the purpose of the data analysis to obtain analysis result data; generating a prompt for inputting a language model that instructs interpretation of the analysis result data based on gene-related information associated with the analysis result data and a purpose of the data analysis; and obtaining an interpretation result of the analysis result data based on the prompt and the language model. An information processing method comprising: [Item 20] a step of analyzing the data related to genetic information according to the purpose of the data analysis and obtaining analysis result data; generating a prompt for input of a language model that instructs the user to interpret the analysis result data based on gene-related information associated with the analysis result data and a purpose of the data analysis; obtaining an interpretation of the analysis result data based on the prompt and the language model; A computer program for causing a computer to execute the above. [Explanation of symbols]

[0095] 1 Processing section 10 Data input section (input section) 20 Dataset Search Section 30 Meta-Analysis Section 40 Interpretation Generation Unit 50 Data output section (output section) 201 Search Query Generation Unit 202 Dataset Search Section 203 Dataset Evaluation Section 204 Dataset Selection Section 205 Public Dataset DB (Third Database) 301 Analytical Data Processing Department 302 Meta-Analysis Applications 303 Language Model Input Data Processing Unit 401 Prompt generation unit for interpretation generation 402 Language Model Interpretation and Generation Unit 403 User Presentation Data Processing Unit 404 Gene-related Information DB (1st Database) 405 Gene-related Additional Information Search Section 406 Gene-related additional information DB (2nd database) 600 Computer equipment 602 Input Interface 603 Display device 604 Communication equipment 605 Main storage 606 External storage device 607 Bus

Claims

1. Analyze the genetic information data according to the purpose of the data analysis to obtain analysis result data; a processing unit that generates a prompt for inputting a language model that instructs interpretation of the analysis result data based on gene-related information associated with the analysis result data and a purpose of the data analysis, and obtains an interpretation result of the analysis result data based on the prompt and the language model; An information processing device comprising:

2. The processing unit acquires a template for creating the prompt, the template including a first variable for storing the analysis result data, a second variable for storing the gene-related information related to the analysis result data, and a third variable for storing a purpose of the data analysis, and generates the prompt by storing the analysis result data in the first variable of the template, the gene information in the second variable, and the purpose of the data analysis in the third variable. The information processing device according to claim 1 .

3. the processing unit evaluates the quality of the interpretation result, regenerates the prompt in accordance with the evaluation result, and obtains an interpretation result of the analysis result data based on the regenerated prompt and the language model. The information processing device according to claim 1 .

4. an output unit that outputs the interpretation result; an input unit for inputting information, the processing unit acquires additional information for the prompt from the input unit, and regenerates the prompt further based on the acquired additional information. The information processing device according to claim 3 .

5. The additional information obtained includes a user question regarding the interpretation result; the processing unit obtains an answer to the question together with an interpretation result of the analysis result data based on the regenerated prompt and the language model. The information processing device according to claim 4 .

6. The acquired additional information includes interpretation results previously acquired for data analysis with the same purpose as the purpose of the data analysis; the processing unit acquires information describing a difference between the interpretation result of the analysis result data and the previously acquired interpretation result, based on the regenerated prompt and the language model. The information processing device according to claim 4 .

7. the processing unit searches a first database storing a plurality of pieces of gene-related information based on the analysis result data, thereby acquiring the gene-related information related to the analysis result data; The information processing device according to claim 1 .

8. the processing unit searches a second database that stores a plurality of pieces of other gene-related information different from the gene-related information based on the analysis result data, thereby acquiring the other gene-related information related to the analysis result data; generating the prompt further based on the acquired additional gene-related information; The information processing device according to claim 7 .

9. The analysis result data includes a plurality of genes according to the purpose of the data analysis. The information processing device according to claim 1 .

10. The gene-related information includes information on genes that are putative targets of the genes included in the analysis result data. The information processing device according to claim 9 .

11. the processing unit generates the prompt based on other gene-related information related to the analysis result data that is different from the gene-related information; The other gene-related information is information on a paper related to the gene included in the analysis result data. The information processing device according to claim 10.

12. The processing unit performs the data analysis by executing an analysis algorithm according to the purpose of the data analysis. The information processing device according to claim 1 .

13. The data analysis is a meta-analysis. The information processing device according to claim 12.

14. The data related to the genetic information is a dataset including measurement data related to genes for a plurality of specimens. The processing device of claim 1 .

15. the processing unit acquires the data related to the genetic information by searching a third database that stores a plurality of data related to the genetic information, based on at least one of a purpose of the data analysis and a search condition for the data related to the genetic information. The information processing device according to claim 1 .

16. the search conditions include at least one of a condition specifying a database to be used for the search and a condition specifying a measuring device used to measure the genetic information; the processing unit searches a database that satisfies the search conditions, or searches for data in which the genetic information is measured by the measurement device. The information processing device according to claim 15.

17. the data related to the genetic information and each of the data in the third database are datasets including measurement data related to genes for a plurality of specimens, the processing unit acquires a plurality of data sets from the third database; selecting a plurality of datasets from the plurality of acquired datasets based on the number of specimens and the genes to be measured in each of the plurality of acquired datasets, and analyzing the selected plurality of datasets to obtain the analysis result data; The information processing device according to claim 15.

18. an output unit that outputs information on the selected plurality of data sets; an input unit for inputting information, the processing unit acquires, from the input unit, selection information for selecting a part of the selected plurality of data sets or addition instruction information for instructing addition of a new data set; updating a set of the selected plurality of datasets based on the selection information or the additional instruction information, and analyzing the datasets of the updated set to obtain the analysis result data; The information processing device according to claim 17.

19. Analyze the genetic information data according to the purpose of the data analysis to obtain analysis result data; generating a prompt for inputting a language model that instructs interpretation of the analysis result data based on gene-related information associated with the analysis result data and a purpose of the data analysis; and obtaining an interpretation result of the analysis result data based on the prompt and the language model. An information processing method comprising:

20. a step of analyzing the data related to genetic information according to the purpose of the data analysis and obtaining analysis result data; generating a prompt for input of a language model that instructs the user to interpret the analysis result data based on gene-related information associated with the analysis result data and a purpose of the data analysis; obtaining an interpretation of the analysis result data based on the prompt and the language model; A computer program for causing a computer to execute the above.

Images