LLM-based proteome data analysis and data mining system
Through a system based on large language model, natural language descriptions are analyzed and proteome data analysis is solved, and data analysis is achieved without programming skills for people in biology background, and efficient proteome data analysis and data mining are achieved without programming .
Patent Information
- Application Number
- CN202410935536.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-12
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-12
AI Technical Summary
It is difficult to effectively analyze proteome data in the prior art, especially due to the lack of programming skills of people with biology backgrounds, proteome data processing and analysis are complex, making it difficult to achieve efficient data mining.
Using a system based on the large language model (LLM), the natural language understanding module receives user input, analyzes user intentions and converts them into standard analysis mode, combines data storage and analysis modules to perform proteome data analysis, and provides report display to achieve programming-free Complex data analysis.
This reduces the complexity of proteomic data analysis. Ordinary users can complete complex proteomic data analysis and data mining through natural language description, breaking the barriers between humans and proteomic data analysis and improving analysis efficiency.
Smart Images

Figure CN118942556A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of artificial intelligence technology and proteomics technology, and specifically relates to a proteomics data analysis and data mining system based on LLM. Background Art
[0002] Proteomics is the study of all proteins in an organism, cell or tissue, collectively known as the proteome. Proteomics focuses on understanding the properties of proteins, such as structure, function, interaction, expression level and post-translational modification, and how these properties change under different conditions, including changes in health and disease. The importance of proteomics research lies in the fact that although the genome provides the basis of genetic information, proteins are the entities that perform most biological functions. The expression level, activity and interaction of proteins may vary depending on cell type, developmental stage, physiological state and environmental factors, so proteomics helps to reveal how organisms respond to internal and external stimuli, as well as the molecular mechanisms of disease occurrence. With the advancement of technology, especially the development of mass spectrometry technology and bioinformatics, proteomics research has been able to achieve large-scale, high-throughput analysis, which has greatly promoted the understanding and application of life sciences, especially in precision medicine, drug discovery and biomarker identification. Analyzing proteomics data is a multidisciplinary task that requires a series of professional skills and knowledge background. People engaged in proteomic analysis are basically from a biological background and have a solid biological background, but they lack programming skills, especially the ability to use languages such as R, Python or Perl for data processing and scripting. Summary of the invention
[0003] Purpose of the invention: To solve the problems existing in the prior art, the present invention provides a proteomic data analysis and data mining system based on LLM.
[0004] Technical solution: In order to achieve the above-mentioned invention object, the present invention adopts the following technical solution:
[0005] A proteomic data analysis and data mining system based on LLM, comprising a user input module, a natural language understanding module LLM, a data storage module, a proteomic analysis module and a report display module;
[0006] (1) The user input module is responsible for receiving and obtaining the natural language description of the user input;
[0007] (2) The natural language understanding module LLM is responsible for receiving the natural language instructions of the user input module, extracting the key entities, relationships, and operation intentions, rewriting them into standard analysis modes through the LLM model, and then feeding them back to the user. The user confirms the correct content and analyzes the user's intentions. The user's intentions are divided into three types: new sample proteome analysis, historical sample data query, and historical multi-sample joint analysis. The intentions are rewritten into corresponding analysis statements and SQL statements and sent to the downstream modules;
[0008] (3) a data storage module, receiving the SQL statement converted from the historical sample data query and the historical multi-sample joint analysis intent in (2), querying the historical sample data, and further forwarding the analysis statement in the historical multi-sample joint analysis intent in (2) to the proteomic analysis module;
[0009] (4) The proteome analysis module is responsible for receiving the analysis statement of the new sample proteome analysis intention in (2) and performing the analysis of the new sample proteome; receiving the historical sample data and the analysis statement in the historical multi-sample joint analysis intention queried by the data storage module in (3) to perform the historical multi-sample joint analysis;
[0010] (5) The report display module is responsible for receiving the results generated by the proteomic analysis module in (4) and presenting them to the user in an easy-to-understand form; and receiving the results of the historical sample data query of the data storage module in (3) and presenting them to the user in an easy-to-understand form.
[0011] As a specific implementation scheme, in the user input module, the natural language description refers to the text combined with the target analysis, including the natural language description of new sample analysis, historical sample query and historical multi-project joint analysis;
[0012] Specifically, when analyzing new samples, the natural language description includes the new sample matrix and the content to be analyzed; when querying historical samples, the natural language description includes the description of previous analysis projects; when jointly analyzing historical multiple projects, the natural language description includes the description of previous analysis projects and the content to be jointly analyzed;
[0013] More specifically, when analyzing a new sample, the natural language description includes the protein matrix that the user wants to analyze, a natural language description of the protein matrix, a grouping description of the matrix, and a description of the analysis mode, wherein the description of the analysis mode includes PCA analysis, heat map analysis, data difference comparison analysis, annotation analysis of data difference comparison results, and a description of the visualization requirements of the analysis results.
[0014] When querying historical samples, the natural language description includes a description of the historical analysis project, including a historical project analysis table, and an analysis result of a specific historical project;
[0015] When performing a joint analysis of multiple historical projects, the natural language description includes the project names of the multiple projects, the method of merging the multiple projects, the method of grouping, the analysis mode, including PCA analysis, heat map analysis, data difference comparison analysis, annotation analysis of the data difference comparison results, and a description of the visualization requirements for the analysis results.
[0016] As a specific implementation scheme, in the natural language understanding module LLM, the LLM model is rewritten into a standard analysis mode and then fed back to the user. After the user receives the feedback, if the intent rewritten by the LLM model is consistent with the user's intent, Figure 1 If the intention rewritten by the LLM model is inconsistent with the user's intention, the user rewrites the analysis model and submits it. The LLM model obtains the analysis model rewritten by the user, extracts the key entities, relationships, and operation intentions, rewrites them into a standard analysis model through the LLM model, and feeds them back to the user. This cycle continues until the user confirms the intention. Figure 1 After that, the LLM model analyzes the user intention.
[0017] As a specific implementation scheme, in the natural language understanding module LLM:
[0018] After receiving the natural language instruction of the user input module, the key entities, relationships and operation intentions are extracted. If these entities contain the description of the analysis matrix and the grouping information of the samples, the mapping relationship between the row and column labels and the sample characteristics, and the operation is to perform data dimension analysis, perform difference comparison analysis, annotate the data difference comparison results, and visualize the analysis results, then the user intention is parsed into a new sample proteome analysis, and the intention sequence and the front and back dependencies are further re-deconstructed, organized into a standard analysis mode with front and back dependencies and fed back to the user;
[0019] After receiving the natural language instruction from the user input module, the key entities, relationships and operation intentions are extracted. If these entities include historical project names, project query fields, and how to display the query results, the user intention is parsed into historical sample data queries, and these intentions are further deconstructed and fed back to the user;
[0020] After receiving the natural language instructions from the user input module, the key entities, relationships and operation intentions are extracted. If these entities contain multiple historical project names, the fields corresponding to the project names are regrouped, the operations are data dimension analysis, data difference analysis, and the results of the difference analysis are annotated. How to visualize all analysis results, the user intention is parsed into a historical multi-sample joint analysis, and these intentions are further deconstructed and fed back to the user.
[0021] As a specific implementation scheme, in the natural language understanding module LLM:
[0022] When the analysis intention is proteomic analysis of a new sample, the intention is rewritten into a corresponding analysis statement, and then the analysis statement is transmitted to the proteomic analysis module;
[0023] When the analysis intention is to query historical sample data, the analysis intention is converted into a SQL statement and sent to the data storage module;
[0024] When the analysis intention is a historical multi-sample joint analysis, the data used in the multiple projects involved are first converted into SQL statements and sent to the data storage module together with the analysis statements in the analysis intention.
[0025] As a specific implementation, in the proteome analysis module:
[0026] The analysis of the proteome of the new sample includes PCA analysis, difference comparison analysis, and difference comparison result annotation; the analysis results are structured and sent to the data storage module for structured storage; and the analysis results are further sent to the report display module for data display;
[0027] The historical multi-sample joint analysis uses historical sample data and analysis statements in the historical multi-sample joint analysis intent to analyze the proteome, including PCA analysis, difference comparison analysis, and difference comparison result annotation; the analysis results are structured and sent to the data storage module for structured storage, and the analysis results are further sent to the report display module for data display.
[0028] As a specific implementation scheme, in the report presentation module, the easily understandable form includes graphs and tables.
[0029] As a specific implementation scheme, the report display module also includes a download function.
[0030] LLM (Large Language Models) refers to a type of artificial intelligence language model with a very large number of parameters that is trained using deep learning technology. The development of LLM represents an important milestone in the field of natural language processing. LLM can not only understand human language, but also generate computer code based on human language, which can assist people without programming skills to write computing programs and complete computing tasks. The capabilities of the LLM large model happen to assist biological workers engaged in proteomic data analysis and help them handle complex programming tasks. To this end, the present invention proposes a proteomic data analysis and data mining system based on LLM, which can complete complex proteomic analysis tasks through natural language descriptions to solve the above problems.
[0031] Beneficial effects: Compared with the prior art, the present invention provides a proteomic data analysis and data mining system based on LLM, which can break the barriers between humans and complex proteomic data analysis and reduce the complexity of analysis. That is, ordinary users can complete complex proteomic data analysis and data mining through natural language descriptions and obtain desired results. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] Figure 1 It is a schematic diagram of the process of the present invention;
[0033] Figure 2 This is the analysis function possessed by the proteome analysis module of the present invention.
[0034] Figure 3 The protein expression matrix.
[0035] Figure 4 The analysis result graph (PCA) in the report display module.
[0036] Figure 5 Displays analysis results in the report module (heat map analysis). DETAILED DESCRIPTION
[0037] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are only embodiments of a part of the present application, rather than all embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work should fall within the scope of protection of the present application.
[0038] Example 1
[0039] A proteomic data analysis and data mining system based on LLM is mainly used to assist workers engaged in biological research in proteomic data analysis and mining. The technical solution is as follows:
[0040] The system includes a user input module, a natural language understanding module LLM, a data storage module, a proteome analysis module and a report display module; the main functions and usage methods are as follows:
[0041] 1. Enter the content you want to analyze in the user input module, including analysis of new samples, query of historical samples, and historical multi-project joint analysis. For new sample analysis, enter the new sample matrix and the content you want to analyze. For historical sample query, enter the description of historical sample analysis content. For historical multi-project joint analysis, enter the description of previous analysis projects and the content you want to jointly analyze.
[0042] When analyzing a new sample, the natural language description includes the protein matrix that the user wants to analyze, the natural language description of the protein matrix, the grouping description of the matrix, and the description of the analysis mode, wherein the description of the analysis mode includes PCA analysis, heat map analysis, data difference comparison analysis, annotation analysis of data difference comparison results, and a description of the visualization requirements of the analysis results.
[0043] When querying historical samples, the natural language description includes a description of the historical analysis project, including a historical project analysis table, and an analysis result of a specific historical project;
[0044] When performing a joint analysis of multiple historical projects, the natural language description includes the project names of the multiple projects, the method of merging the multiple projects, the method of grouping, the analysis mode, including PCA analysis, heat map analysis, data difference comparison analysis, annotation analysis of the data difference comparison results, and a description of the visualization requirements for the analysis results.
[0045] 2. The natural language understanding module LLM large model vectorizes the user input obtained. The vectorization algorithm is built based on the BGE (BAAI General Embedding) algorithm built on BERT. The constructed vectorized text and the vector database are retrieved in context based on vector similarity. The retrieved text contains three intents: new sample analysis intent, historical sample query intent, and historical multi-sample joint analysis intent.
[0046] After receiving the natural language instructions from the user input module, the key entities, relationships and operation intentions are extracted. If these entities contain the description of the analysis matrix and sample grouping information, the mapping relationship between row and column labels and sample features, and the operation is to perform data dimension analysis, perform difference comparison analysis, annotate the data difference comparison results, and visualize the analysis results, then the user intention is parsed into a new sample proteome analysis, and the intention sequence and front-end dependencies are further re-deconstructed and organized into a standard analysis mode with front-end dependencies to be fed back to the user;
[0047] After receiving the natural language instructions from the user input module, the key entities, relationships and operation intentions are extracted. If these entities contain historical project names, project query fields, and how to display query results, the user intention is parsed into historical sample data queries, and these intentions are further deconstructed and fed back to the user;
[0048] After receiving the natural language instructions from the user input module, the key entities, relationships and operation intentions are extracted. If these entities contain multiple historical project names, the fields corresponding to the project names, the form of regrouping, the operations are data dimension analysis, data difference analysis, annotations on the results of the difference analysis, and how to visualize all analysis results, the user intention is parsed into a historical multi-sample joint analysis, and these intentions are further deconstructed and fed back to the user.
[0049] 3. After the user receives the modified analysis intent of the LLM model, if the intent is consistent with the user's Figure 1 The user inputs the form of affirmative intent together with the modified intent of the LLM model. Figure 1 If the intent is inconsistent with the user's intent, the user needs to re-describe the inconsistent part of the intent after the LLM model is modified, and then send it to the LLM model together with the modified intent. Figure 1 Same as the large model sent to LLM.
[0050] 4. After the LLM model receives the user's confirmation response to the intent, it proceeds to the next step based on different analysis intentions:
[0051] Furthermore, if the analysis intent is a new sample proteome analysis, the intent is converted into an analysis statement and sent to the proteome analysis module.
[0052] Furthermore, if the analysis intention is a historical sample query, the intention is converted into a SQL statement and sent to the data storage module.
[0053] Furthermore, if the analysis intention is a historical multi-project joint analysis, the intention is converted into a SQL statement and an analysis statement and then sent to the data storage module.
[0054] 5. The data storage module receives queries for historical projects, and directly sends the query results to the report display module; after receiving queries for historical multi-project joint analysis, the query results are sent to the proteomic analysis module for multi-project joint analysis.
[0055] 6. The analysis received by the proteomic analysis module includes the proteomic analysis of new samples and the joint analysis of historical multiple samples.
[0056] The analysis of the proteome of the new sample includes PCA analysis, differential comparison analysis, and annotation of differential comparison results; the analysis results are structured and sent to the data storage module for structured storage; the analysis results are further sent to the report display module for data display;
[0057] Historical multi-sample joint analysis uses historical sample data and analysis statements in historical multi-sample joint analysis intent to analyze the proteome, including PCA analysis, differential comparison analysis, and differential comparison result annotation; the analysis results are structured and sent to the data storage module for structured storage, and further the analysis results are sent to the report display module for data display.
[0058] A more specific method is as follows:
[0059] The new sample proteome analysis includes:
[0060] Normalize protein data to mean variance: x i is the expression level of one of the proteins, μ i is the mean of one protein in all samples, σ i is the standard deviation of one protein in all samples, x′ i is the normalized protein expression level of one of the proteins;
[0061] PCA analysis based on the scikit-learn module in Python;
[0062] Heatmap analysis based on the ComplexHeatmap module in R;
[0063] The difference analysis was performed using the t test. When the variance homogeneity test of two proteins in the two groups was unequal, the When the two proteins are equal in the two groups, the homogeneity test is used. in, is the average value of a protein in group A, is the average value of a protein in group B, is the variance of a protein in group A, is the variance of a protein in group B, n A is the number of samples A, n B is the number of samples B.
[0064] GO / KEGG enrichment analysis, enrichment analysis N is the number of proteins with GO / KEGG annotation information among all proteins, n is the number of differential proteins in N, M is the number of proteins annotated to a certain GO / KEGG entry among all proteins, and x is the number of differential proteins annotated to a certain GO / KEGG entry.
[0065] Annotation of differential protein drug database: annotate the obtained differential proteins with the own drug database.
[0066] Furthermore, all analysis results are stored in the data storage module as structured data.
[0067] 7. The report display module displays the data obtained by the proteomics analysis module in the form of doc, html or pdf.
[0068] Example 2
[0069] like Figure 1-2 As shown, a proteomic data analysis and data mining system based on LLM includes a user input module, a natural language understanding module LLM, a proteomic analysis module, a data storage module, and a report display module.
[0070] Step 1: In the user input module, the user uses natural language to describe the analysis he wants to do in detail and clearly. For example, if the user wants to perform proteomic analysis based on a new project, Figure 3 In the example of protein expression matrix shown, the user first transfers his analysis data to the system in the form of a file, and then describes the data: "columns represent features, rows represent samples, column labels represent protein names, grouping labels are in the second column, features start from the third column, samples start from the second row, group A is the control group, group B is the experimental group, PCA analysis is required, heat map analysis is required, differential analysis is required, GO / KEGG enrichment analysis is required after differential analysis, and drug annotation analysis is required".
[0071] Step 2: After receiving the natural language from the user input module, the natural language understanding module LLM parses the user description and finds that it includes a description of the matrix information and the analysis content. It determines that the user's intention is to analyze the proteome of a new sample and rewrites the intention as "protein matrix (row: sample; starting from the second row, column: feature, starting from the third column); grouped as [A, B], located in the second column; analysis content is [PCA, Heatmap, Diff, GO, KEGG, Drug]".
[0072] Step 3: After the user input module receives the intent rewritten by the natural language understanding module LLM in step 2, it is consistent with its own description, and adds affirmative information after the rewritten intent. At this time, the intent becomes: "Protein matrix (row: sample; starting from the second row, column: feature, starting from the third column); grouped as [A, B], located in the second column; analysis content is [PCA, Heatmap, Diff, GO, KEGG, Drug]; no problem", and the user confirms and sends the rewritten intent to the natural language understanding module LLM. The rewritten intent does not meet the user's analysis intent. The user rewrites the analyzed intent, such as modifying the description after grouping: "Protein matrix (row: sample; starting from the second row, column: feature, starting from the third column); grouped as [A, B], located in the second column; analysis content is [PCA, Heatmap, Diff, GO, KEGG, Drug]; grouping error, grouping is [A, C], others are correct", and sends the rewritten intent to the natural language understanding module LLM.
[0073] Step 4: After receiving the user confirmation intent in step 3, the natural language understanding module LLM converts the analysis intent into an analysis statement: "feature = col[3:]; sample = row[2:]; group = col[2]; groups = [A, B]; analysis = [PCA, Heatmap, Diff, GO, KEGG, Drug]" and sends it to the proteomics analysis module; if the received intent is a modification of the intent in step 3, the intent is further rewritten as "protein matrix (row: sample; starting from the second row, column: feature, starting from the third column); grouped as [A, C], located in the second column; analysis content is [PCA, Heatmap, Diff, GO, KEGG, Drug]", and returns to step 3, and repeats this process until the user confirms that the intent is correct. In this example, the user only modified the intent once and confirmed that there was no problem. Finally, the intent after grouping was only modified to "Protein matrix (row: sample; starting from the second row, column: feature, starting from the third column); grouped as [A, C], located in the second column; analysis content is [PCA, Heatmap, Diff, GO, KEGG, Drug]; no problem"; after receiving the intent, LLM rewrote it into an analysis statement: "feature = col[3:]; sample = row[2:]; group = col[2]; groups = [A, C]; analysis = [PCA, Heatmap, Diff, GO, KEGG, Drug]" and sent it to the proteomics analysis module.
[0074] Step 5: The proteomic analysis module receives the new sample proteomic analysis statement "feature = col[3:]; sample = row[2:]; group = col[2]; groups = [A, C]; analysis = [PCA, Heatmap, Diff, GO, KEGG, Drug]" sent by the natural language understanding module LLM, verifies the protein data and the analysis statement, splits the analysis statement by ";", reads the keywords "feature, sample, group, groups and analysis", verifies the assignment information corresponding to the keywords "feature, sample, group, groups" with the protein matrix, and then determines whether the analysis corresponding to "analysis" is included in the following. Figure 2 If the sample is found to be inconsistent with the sample description, such as the protein matrix is grouped as A and C, but described as A and B, or the analysis content in "analysis" is inconsistent with the sample description, the analysis result is sent to the report display module after the analysis is completed. Figure 2 If there is a conflict in the analysis module, the analysis is terminated and the cause of the error is sent to the report display module.
[0075] Step 6: The data storage module receives the formatted protein analysis results in step 5, including protein project information, group analysis, analysis content result path, etc. and stores them in the database.
[0076] Step 7: The report display module receives the analysis results of the proteomic analysis module in step 5 and displays them to the user in doc, html and pdf format for download. Some analysis results are shown in Figure 4 (PCA), Figure 5 (Heat map analysis).
[0077] The above description is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.
Claims
1. A proteomic data analysis and data mining system based on LLM, characterized in that: It includes user input module, natural language understanding module LLM, data storage module, proteome analysis module and report display module; (1) The user input module is responsible for receiving and obtaining the natural language description of the user input; (2) The natural language understanding module LLM is responsible for receiving the natural language instructions of the user input module, extracting the key entities, relationships, and operation intentions, rewriting them into standard analysis modes through the LLM model, and then feeding them back to the user. The user confirms the correct content and analyzes the user's intentions. The user's intentions are divided into three types: new sample proteome analysis, historical sample data query, and historical multi-sample joint analysis. The intentions are rewritten into corresponding analysis statements and SQL statements and sent to the downstream modules; (3) a data storage module, receiving the SQL statement converted from the historical sample data query and the historical multi-sample joint analysis intent in (2), querying the historical sample data, and further forwarding the analysis statement in the historical multi-sample joint analysis intent in (2) to the proteomic analysis module; (4) The proteome analysis module is responsible for receiving the analysis statement of the new sample proteome analysis intention in (2) and performing the analysis of the new sample proteome; receiving the historical sample data and the analysis statement in the historical multi-sample joint analysis intention queried by the data storage module in (3) to perform the historical multi-sample joint analysis; (5) The report display module is responsible for receiving the results generated by the proteomic analysis module in (4) and presenting them to the user in an easy-to-understand form; and receiving the results of the historical sample data query of the data storage module in (3) and presenting them to the user in an easy-to-understand form.
2. The LLM-based proteomic data analysis and data mining system according to claim 1, characterized in that: In the user input module, the natural language description refers to the text combined with the target analysis, including the natural language description of new sample analysis, historical sample query and historical multi-project joint analysis; Specifically, when analyzing new samples, the natural language description includes the new sample matrix and the content to be analyzed; when querying historical samples, the natural language description includes the description of previous analysis projects; when jointly analyzing historical multiple projects, the natural language description includes the description of previous analysis projects and the content to be jointly analyzed; More specifically, when analyzing a new sample, the natural language description includes the protein matrix that the user wants to analyze, a natural language description of the protein matrix, a grouping description of the matrix, and a description of the analysis mode, wherein the description of the analysis mode includes PCA analysis, heat map analysis, data difference comparison analysis, annotation analysis of data difference comparison results, and a description of the visualization requirements of the analysis results. When querying historical samples, the natural language description includes a description of the historical analysis project, including a historical project analysis table, and an analysis result of a specific historical project; When performing a joint analysis of multiple historical projects, the natural language description includes the project names of the multiple projects, the method of merging the multiple projects, the method of grouping, the analysis mode, including PCA analysis, heat map analysis, data difference comparison analysis, annotation analysis of the data difference comparison results, and a description of the visualization requirements for the analysis results.
3. The LLM-based proteomic data analysis and data mining system according to claim 1, characterized in that: In the natural language understanding module LLM, the LLM model is rewritten into a standard analysis mode and then fed back to the user. After the user receives the feedback, if the intention rewritten by the LLM model is consistent with the user's intention, the user only needs to describe it with an affirmative statement and then submit it; If the intent rewritten by the LLM model is inconsistent with the user's intent, the user rewrites the analysis model and then submits it. The LLM model obtains the analysis model rewritten by the user, extracts the key entities, relationships, and operation intentions, and rewrites it into a standard analysis model through the LLM model and feeds it back to the user. This cycle continues until the user confirms that the intent is consistent, and the LLM model then parses the user's intent again.
4. The LLM-based proteomic data analysis and data mining system according to claim 1, characterized in that: In the natural language understanding module LLM: After receiving the natural language instruction of the user input module, the key entities, relationships and operation intentions are extracted. If these entities contain the description of the analysis matrix and the grouping information of the samples, the mapping relationship between the row and column labels and the sample characteristics, and the operation is to perform data dimension analysis, perform difference comparison analysis, annotate the data difference comparison results, and visualize the analysis results, then the user intention is parsed into a new sample proteome analysis, and the intention sequence and the front and back dependencies are further re-deconstructed, organized into a standard analysis mode with front and back dependencies and fed back to the user; After receiving the natural language instruction from the user input module, the key entities, relationships and operation intentions are extracted. If these entities include historical project names, project query fields, and how to display the query results, the user intention is parsed into historical sample data queries, and these intentions are further deconstructed and fed back to the user; After receiving the natural language instructions from the user input module, the key entities, relationships and operation intentions are extracted. If these entities contain multiple historical project names, the fields corresponding to the project names are regrouped, the operations are data dimension analysis, data difference analysis, and the results of the difference analysis are annotated. How to visualize all analysis results, the user intention is parsed into a historical multi-sample joint analysis, and these intentions are further deconstructed and fed back to the user.
5. The LLM-based proteomic data analysis and data mining system according to claim 1, characterized in that: In the natural language understanding module LLM: When the analysis intention is proteomic analysis of a new sample, the intention is rewritten into a corresponding analysis statement, and then the analysis statement is transmitted to the proteomic analysis module; When the analysis intention is to query historical sample data, the analysis intention is converted into a SQL statement and sent to the data storage module; When the analysis intention is a historical multi-sample joint analysis, the data used in the multiple projects involved are first converted into SQL statements and sent to the data storage module together with the analysis statements in the analysis intention.
6. The LLM-based proteomic data analysis and data mining system according to claim 1, characterized in that: In the proteome analysis module: The analysis of the proteome of the new sample includes PCA analysis, difference comparison analysis, and difference comparison result annotation; the analysis results are structured and sent to the data storage module for structured storage; and the analysis results are further sent to the report display module for data display; The historical multi-sample joint analysis uses historical sample data and analysis statements in the historical multi-sample joint analysis intent to analyze the proteome, including PCA analysis, difference comparison analysis, and difference comparison result annotation; the analysis results are structured and sent to the data storage module for structured storage, and the analysis results are further sent to the report display module for data display.
7. The LLM-based proteomic data analysis and data mining system according to claim 1, characterized in that: In the report presentation module, the easily understandable form includes graphs and tables.
8. The LLM-based proteomic data analysis and data mining system according to claim 1, characterized in that: The report display module also includes a download function.
Citation Information
Patent Citations
Intelligent data analysis method and device based on large language model
CN117555986A
Data analysis method and system based on natural language, electronic equipment and storage medium
CN118132587A
Engineering of antigen-binding proteins
WO2024079280A1
Cited By
Health data prediction method and device based on protein expression data, equipment and storage medium
CN120808895A
