Risk prediction variant site screening and evaluation system and method based on artificial intelligence

Abstract

The present invention relates to the fields of bioinformatics, gene detection and medicine, and in particular to a risk prediction variant site screening evaluation system and method based on artificial intelligence, the system comprising: a literature retrieval module, an evaluation index extraction module, a scoring module, a variant site screening module and a user interaction module. The system supports the functions of automated literature retrieval, evaluation index extraction, site scoring, variant site screening, etc., which can not only effectively improve the efficiency and accuracy of variant site screening, but also effectively screen out high-quality variant sites, so as to provide important support for disease risk prediction, personalized medicine and clinical gene testing.

Description

Technical Field

[0001] The present invention relates to the fields of bioinformatics, gene detection and medicine, and in particular to a risk prediction variant site screening and evaluation system and method based on a large model, and in particular to a risk prediction variant site screening and evaluation system and method based on artificial intelligence. Background Art

[0002] In the context of precision medicine, the identification, evaluation and screening of genetic variant sites are of great significance for disease risk prediction and personalized medicine. Traditional variant site screening methods usually rely on manual literature review and expert experience, which have problems such as low efficiency, difficulty in large-scale article review, limited number of manually reviewed literature and not necessarily comprehensive, time-consuming and labor-intensive, and easily affected by subjective factors, making it difficult to systematize and scale.

[0003] Therefore, there is an urgent need for an artificial intelligence-based risk prediction variant site screening and evaluation system and method that can efficiently and accurately screen and evaluate variant sites to assist in clinical genetic testing and risk prediction. Summary of the invention

[0004] Based on this, it is necessary to provide an artificial intelligence-based risk prediction variant site screening and evaluation system and method to address the above-mentioned technical issues.

[0005] According to a first aspect of the present invention, there is provided an artificial intelligence-based risk prediction variant site screening and evaluation system, comprising a literature retrieval module, an evaluation index extraction module, a scoring module, a variant site screening module and a user interaction module; the literature retrieval module is used to receive and respond to a target risk factor keyword query request sent by the user interaction module, call the literature database API through the LLM large language model, search for a target literature data list related to the target risk factor keyword, and send the target literature data list to the user interaction module; the evaluation index extraction module is used to extract all target literature data containing variant site evaluation index fields and key indicator fields from the target literature data list, based on the key indicator fields, The gene variant site omics information fields related to the variant site are obtained from the omics database, and the key indicator fields and the gene variant site omics information fields are structured to obtain all field information of the variant site stored in CSV format data, and all field information of the variant site is sent to the user interaction module; the scoring module is used to construct a variant site evaluation system according to preset evaluation indicators, and assign corresponding scoring standards to each evaluation indicator, and according to each evaluation indicator and the corresponding scoring standard, all target literature data are integrated to calculate each evaluation indicator score and comprehensive score corresponding to each variant site; wherein a mapping relationship is formed between the preset evaluation indicator and the evaluation indicator field; the variant site screening The selection module is used to sort the comprehensive scores of each variant site from high to low, perform site rating on the variant site according to the preset variant site rating rules, obtain the variant site rating results of each variant site, and send the scores of each evaluation indicator, the comprehensive score and the corresponding variant site rating results of each variant site to the user interaction module; receive and respond to the screening condition query request or the target variant site query request sent by the user interaction module, perform dynamic site screening on the variant site, obtain the condition screening results corresponding to the screening condition query request or the target screening results corresponding to the target variant site query request, and send the obtained condition screening results or target screening results to the user interaction module; and process the condition screening results and / or target screening results The results are dynamically combined and named to create a risk site panel; the user interaction module is used to obtain the target risk factor keyword query request, screening condition query request and target variant site query request input by the user, send the target risk factor keyword query request to the literature retrieval module, send the screening condition query request and the target variant site query request to the variant site screening module, store, search, display and export the target literature data list, all field information of the variant site, each evaluation indicator score corresponding to each variant site, the comprehensive score, the corresponding variant site rating result, the condition screening result and the target screening result, and display and manage the relevant information corresponding to the created risk site panel.

[0006] Optionally, the document retrieval module includes: a first response submodule, which is used to receive and respond to the target risk factor keyword query request sent by the user interaction module; a user input preprocessing submodule, which is used to remove stop words and extract stems from the target risk factor keywords contained in the target risk factor keyword query request to obtain valid keywords; an expansion and parsing submodule, which is used to parse the valid keywords through the LLM large language model to obtain parsed medical term keywords and extended synonyms related to the valid keywords; a query string generation submodule, which is used to generate a query string for interacting with the document database based on the parsed medical term keywords and extended synonyms; a screening submodule, which is used to call the document database through the document database API, and based on the query string, screen the titles and abstracts of all document data in the document database, or use a character similarity algorithm to calculate the similarity between the title or abstract of the document data and the query string, and screen out document data related to the target risk factor keywords; a first formatting processing submodule, which is used to format the screened document data related to the target risk factor keywords to obtain target document data stored in JSON format, generate a target document data list based on the target document data, and send the target document data list to the user interaction module.

[0007] Optionally, the evaluation index extraction module includes: a literature data preprocessing submodule, which is used to perform data cleaning and standardization processing on all target literature data in the target literature data list to obtain target literature processing data; a first information extraction submodule, which is used to perform information extraction processing on the target literature processing data through a machine learning algorithm to extract evaluation index fields and key indicator fields; the evaluation index fields include a journal influence field, a sample size field, a research population field, an evidence level field, a variant site OR / HR value field and a variant site literature report number field, and the key indicator field includes a gene name field, a RefSNP identifier field, a risk factor field and an effect allele field corresponding to the variant site; a second information extraction submodule, which is used to extract the target literature processing data from the group according to the key indicator field. The gene variation site group information field related to the variation site is obtained from the gene variation site database; the gene variation site group information field includes a gene field, a CDS mutation syntax field, an amino acid replacement field, a chromosome field, an Hg start position field, an Hg end position field, a reference sequence field, an alternative sequence field, a chain field, a genome assembly version field, a position field, a MAF value field and a variation mechanism field; a second structured processing submodule is used to perform structured processing on the key indicator field and the gene variation site group information field to obtain all field information of the variation site stored in CSV format data, and send all field information of the variation site to the user interaction module, wherein all field information of the variation site includes the key indicator field and the gene variation site group information field corresponding to the variation site.

[0008] Optionally, the scoring module includes: an evaluation system construction submodule, which is used to construct a variant site evaluation system according to preset evaluation indicators, and assign corresponding scoring standards to each evaluation indicator; wherein a mapping relationship is formed between the preset evaluation indicators and the evaluation indicator fields, and the evaluation indicators include journal influence, sample size, research population, evidence level, variant site OR / HR value and number of literature reports on the variant site; a scoring calculation submodule, which is used to calculate the score of each evaluation indicator and the comprehensive score corresponding to each variant site based on various evaluation indicators and corresponding scoring standards, and comprehensively integrate all target literature data.

[0009] Optionally, the variant site screening module includes: a site rating submodule, which is used to sort the comprehensive scores of each variant site from high to low, and to perform site rating on the variant sites according to preset variant site rating rules to obtain variant site rating results for each variant site, and to send the scores of each evaluation indicator, the comprehensive score and the corresponding variant site rating results corresponding to each variant site to the user interaction module; a second response submodule, which is used to receive and respond to the screening condition query request or the target variant site query request sent by the user interaction module, perform dynamic site screening on the variant sites, obtain the condition screening results corresponding to the screening condition query request or the target screening results corresponding to the target variant site query request, and send the obtained condition screening results or the target screening results to the user interaction module; a risk site panel creation submodule, which is used to dynamically combine and name the screening results according to user needs to create a risk site panel.

[0010] Optionally, the user interaction module includes: a Panel management submodule, which is used to display, add, delete and update relevant information corresponding to the created risk site Panel, wherein the relevant information corresponding to the risk site Panel includes PanelID, name, number of sites and site information, and the site information includes the score of each evaluation indicator corresponding to the variant site, the comprehensive score, the ectopic site rating result, the target literature data related to the corresponding variant site, the evaluation indicator field related to the corresponding variant site, the key indicator field related to the corresponding variant site and the gene variant site omics information field; a site management submodule, which obtains the target risk factor keyword query request, screening condition query request and target variant site query request input by the user, sends the target risk factor keyword query request to the literature retrieval module, sends the screening condition query request and the target variant site query request to the variant site screening module, stores, searches, displays and exports the target literature data list, all field information of the variant site, the score of each evaluation indicator corresponding to each variant site, the comprehensive score, the corresponding variant site rating result, the condition screening result and the target screening result.

[0011] Optionally, the site management submodule also includes: a variant site submodule, which is used to obtain a target variant site query request input by a user, send the target variant site query request to a variant site screening module, store, search, display and export target screening results, and store, search, display all field information of the variant site; a risk prediction submodule, which is used to obtain a target risk factor keyword query request and screening condition query request input by a user, store, search, display and export a target literature data list, a score for each evaluation indicator corresponding to each variant site, a comprehensive score, the corresponding variant site rating results and condition screening results.

[0012] According to a second aspect of the present invention, a risk prediction variant site screening and evaluation method based on artificial intelligence is provided, the method comprising: obtaining a target risk factor keyword query request, a screening condition query request and a target variant site query request input by a user through a user interaction module, sending the target risk factor keyword query request to a document retrieval module, sending the screening condition query request and the target variant site query request to the variant site screening module; receiving and responding to the target risk factor keyword query request sent by the user interaction module through the document retrieval module, calling the document database API through the LLM large language model, searching for a target document data list related to the target risk factor keyword, and sending the target document data list to the target document screening module. The target literature data list is sent to the user interaction module; the evaluation index fields and key indicator fields of all target literature data containing variant sites are extracted from the target literature data list through the evaluation index extraction module, and the gene variant site omics information fields related to the variant sites are obtained from the omics database based on the key indicator fields, and the key indicator fields and the gene variant site omics information fields are structured to obtain all field information of the variant sites stored in CSV format data, and all field information of the variant sites is sent to the user interaction module; the variant site evaluation system is constructed according to the preset evaluation indicators through the scoring module, and each evaluation indicator is assigned a corresponding scoring standard, and the evaluation criteria are assigned according to each item. Evaluation indicators and corresponding scoring standards are integrated with all target literature data to calculate the scores of each evaluation indicator and the comprehensive score corresponding to each variant site; wherein a mapping relationship is formed between the preset evaluation indicators and the evaluation indicator fields; the comprehensive scores of each variant site are sorted from high to low through the variant site screening module, and the variant sites are rated according to the preset variant site rating rules to obtain the variant site rating results of each variant site, and the scores of each evaluation indicator, the comprehensive score and the corresponding variant site rating results corresponding to each variant site are sent to the user interaction module; the screening condition query request or the target variant site query request sent by the user interaction module is received and responded to, and the variant site is dynamically rated. Site screening, obtaining the condition screening results of the corresponding screening condition query request or the target screening results of the corresponding target variant site query request, and sending the obtained condition screening results or target screening results to the user interaction module; and dynamically combining and naming the condition screening results and / or target screening results to create a risk site panel; storing, searching, displaying and exporting the target literature data list, all field information of the variant site, the score of each evaluation indicator corresponding to each variant site, the comprehensive score, the corresponding variant site rating results, the condition screening results and the target screening results through the user interaction module, and displaying and managing the relevant information corresponding to the created risk site panel.

[0013] According to a third aspect of the present invention, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method when executing the computer program.

[0014] According to a fourth aspect of the present invention, there is provided a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program implements the steps of the method when executed by a processor.

[0015] The advantages and beneficial effects of the present invention are as follows: the risk prediction variant site screening and evaluation system and method based on artificial intelligence provided by the present invention supports the automated implementation of literature retrieval, evaluation index extraction, site scoring, variant site screening and other functions. First, the target risk factor keywords input by the user are parsed through the LLM large language model, and a query string is automatically generated, and the literature database API is called to achieve accurate retrieval of the literature; then, the evaluation indicators and key indicators in the literature are automatically extracted by using a machine learning algorithm, and the gene variant site omics information related to the variant site is obtained from the omics database, the extracted key indicator fields and gene variant site omics information fields are structured and stored, and a comprehensive evaluation is performed through a scoring module, and the comprehensive evaluation index score and comprehensive score of the variant site are calculated according to preset rules and scoring criteria; finally, the variant site screening module is used to sort and screen according to the scoring results, and the user interacts with the system through the user interaction module to view the screening results and export relevant data, and the user can perform personalized combination of Panel sites to construct a risk site panel according to actual needs. This system can not only effectively improve the efficiency and accuracy of variant site screening, but also effectively screen out high-quality variant sites to provide important support for disease risk prediction, personalized medicine and clinical genetic testing. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a system principle block diagram of the present invention.

[0017] Figure 2 This is the system principle block diagram of the literature retrieval module.

[0018] Figure 3 This is the system principle block diagram of the evaluation index extraction module.

[0019] Figure 4 This is the system principle block diagram of the scoring module.

[0020] Figure 5 This is the system principle block diagram of the variant site screening module.

[0021] Figure 6This is the system principle block diagram of the user interaction module.

[0022] Figure 7 It is the overall flow chart of the present invention.

[0023] Figure 8 A schematic diagram of an electronic device. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below through specific implementation methods in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0025] Embodiment 1

[0026] Reference Figure 1-6 , an artificial intelligence-based risk prediction variant site screening and evaluation system 100, the system includes a literature retrieval module 110, an evaluation index extraction module 120, a scoring module 130, a variant site screening module 140 and a user interaction module 150.

[0027] In some embodiments, refer to the attached Figure 2 The document retrieval module 110 is used to receive and respond to the target risk factor keyword query request sent by the user interaction module 150, call the document database API through the LLM large language model, search for the target document data list related to the target risk factor keyword, and send the target document data list to the user interaction module 150.

[0028] Furthermore, the document retrieval module 110 includes: a first response submodule 1110, which is used to receive and respond to the target risk factor keyword query request sent by the user interaction module 150; a user input preprocessing submodule 1120, which is used to remove stop words and extract stems from the target risk factor keywords contained in the target risk factor keyword query request to obtain valid keywords; an expansion and parsing submodule 1130, which is used to parse the valid keywords through the LLM large language model to obtain parsed medical term keywords and extended synonyms related to the valid keywords; a query string generation submodule 1140, which is used to generate a query string corresponding to the document database based on the parsed medical term keywords and extended synonyms. A query string for interaction; a screening submodule 1150, which is used to call the document database through the document database API, and based on the query string, screen the titles and abstracts of all document data in the document database, or use a character similarity algorithm to calculate the similarity between the title or abstract of the document data and the query string, and screen out document data related to the target risk factor keywords; a first formatting processing submodule 1160, which is used to format the screened document data related to the target risk factor keywords, obtain the target document data stored in JSON format, generate a target document data list based on the target document data, and send the target document data list to the user interaction module 150.

[0029] Furthermore, the literature database called by the literature database API includes but is not limited to the PubMed database, the Google Scholar database, the Web of Science database, etc., and can search for relevant literature in the literature database based on the target risk factor keyword query request input by the user. In addition, the LLM large language model adopts natural language processing technology to automatically exclude irrelevant literature, generate literature data related to the target risk factor keywords, and store them in a structured manner in the user interaction module 150.

[0030] Furthermore, the specific implementation process of the document retrieval includes: receiving and responding to the target risk factor keyword query request sent by the user interaction module 150. For example, the target risk factor keyword query request input by the user is "diabetes risk factors", preprocessing the input target risk factor keywords, including removing stop words, stem extraction and other preprocessing to obtain valid keywords, and then using the LLM large language model to parse the valid keywords, and identify the parsed medical term keywords and extended synonyms related to the valid keywords. For example, "diabetes risk factors" may be expanded to "Type 2diabetes risk factors", "Diabetes mellitus risk", etc., and then generate a query string for interacting with the literature database based on the parsed medical term keywords and extended synonyms. It should be noted that this query string may include logical operators (such as AND, OR, NOT, etc.) to improve the accuracy of the retrieval. Then the LLM large language model calls the literature database API, and calls the literature database through the literature database API to screen the title and abstract of each document, or uses a character similarity algorithm to calculate the similarity between the title or abstract of the literature data and the query string, and screens out literature data related to the target risk factor keywords, formats and stores the screened literature data in JSON format, and removes duplicate literature data to avoid repeated processing. Based on the above steps, the complete process from user input to obtaining and screening relevant documents can be automatically realized, and the screened target literature data list is returned in a structured manner and stored in the user interaction module 150.

[0031] Furthermore, the title and abstract of each document are screened, or a character similarity algorithm is used to calculate the similarity between the title or abstract of the document data and the query string, and the document data related to the target risk factor keywords are screened out, including: (1) keyword matching, scanning the title and abstract of the document data to determine whether they contain content that matches the parsed medical term keywords or extended synonyms, where the screening conditions include: whether the target gene, risk factor or effect allele, etc., which matches the parsed medical term keywords or extended synonyms, is explicitly mentioned in the title or abstract. Check whether the title of the literature data contains keywords such as target genes, risk factors or effect alleles. If so, the preliminary screening is passed. On the basis of passing the title screening, further check whether the abstract of the literature data mentions research content related to the target genes, risk factors or effect alleles. If so, the screening is passed. If not, the literature data is filtered out; (2) For literature data that cannot pass the keyword matching, character similarity calculation is required. Use a character similarity algorithm (such as Jaccard similarity) to calculate the similarity between the title or abstract and the parsed medical term keywords or extended synonyms. If the similarity exceeds the preset threshold (such as 70%), the literature data is retained; (3) If both steps (1) and (2) fail, the literature data is filtered out.

[0032] In some embodiments, refer to the attached Figure 3 The evaluation index extraction module 120 is used to extract the evaluation index fields and key indicator fields of all target document data containing variant sites from the target document data list, obtain the gene variant site omics information fields related to the variant sites from the omics database based on the key indicator fields, and perform structured processing on the key indicator fields and the gene variant site omics information fields to obtain all field information of the variant sites stored in CSV format data, and send all field information of the variant sites to the user interaction module 150.

[0033] Furthermore, the evaluation index extraction module 120 includes: a literature data preprocessing submodule 1210, which is used to perform data cleaning and standardization processing on all target literature data in the target literature data list to obtain target literature processing data; a first information extraction submodule 1220, which is used to perform information extraction processing on the target literature processing data through a machine learning algorithm to extract evaluation index fields and key index fields; the evaluation index fields include the journal influence (ImpactFactor) field, the sample size (Sample size) field, the research population (Population) field, the evidence level (Evidence level) field, the variant site OR (odds ratio) / HR (risk ratio) value field and the variant site literature report number field, and the key index fields include the gene name (Gene) field of the corresponding variant site, the RefSNP identifier (RefSNP) field, the risk factor (Risk Factor) field, the effect allele (Effect allele field; the second information extraction submodule 1230 is used to obtain the gene variation site omics information field related to the variation site from the omics database according to the key indicator field; the gene variation site omics information field includes the gene (Gene) field, the CDS mutation syntax (CdsMutation Syntax) field, the amino acid substitution (Amino Acid Substitution) field, the chromosome (Chromosome) field, the Hg start position (Hg Position Start) field, the Hg end position (Hg Position End) field, the reference sequence (Ref Seq) field, the alternative sequence (Alt Seq) field, the chain (Strand) field, the genome assembly version (Assembly) field, the location (Location) field, the MAF value field, the variation mechanism (Mechanism ofVariant) field; a second structured processing submodule 1240 is used to perform structured processing on the key indicator field and the gene variation site omics information field to obtain all field information of the variation site stored in CSV format data, and send all field information of the variation site to the user interaction module 150, where all field information of the variation site includes the key indicator field and the gene variation site omics information field of the corresponding variation site.

[0034] Furthermore, before extracting the evaluation indicators, all target document data in the target document data list are cleaned and standardized, such as title, author, abstract and other fields, to ensure the consistency and integrity of the data.

[0035] Furthermore, the target literature data is processed through machine learning algorithms to extract evaluation index fields and key indicator fields, including: (1) using the BioBERT pre-trained model to automatically identify and extract important information in the target literature data, such as gene name, RefSNP identifier, risk factor, effect allele, study population, sample size, journal name, evidence level, variant site OR (Odds Ratio) / HR (Hazard Ratio) Ratio) value, the number of literature reports on variant sites, and other evaluation indicator fields and key indicator fields; (2) rule matching and regular expressions, for specific evaluation indicators, such as OR / HR value of variant sites, write regular expressions to identify specific patterns (such as: number + CI interval), so as to achieve automatic extraction; (3) relationship extraction, identify the relationship between key indicators and related contexts, and ensure that the extracted key indicators are correctly associated with related content (such as gene name, RefSNP identifier, journal name, evidence level and specific research conclusions); (4) automatically generate journal impact (IF) value based on journal name, which is achieved through the following steps: standardize journal names, clean and normalize journal names to improve matching accuracy; access data sources, access JCR literature database API containing impact factors through subscription; match and search, match the standardized journal names with records in the literature database, and obtain the corresponding journal influence (i.e., impact factor). If no match is found, return the default value or mark "not found".

[0036] Furthermore, the genomic information fields of the gene variant sites related to the variant sites are obtained from the genomic database, wherein the genomic database includes but is not limited to the NCBI-dbSNP database, the Ensembl database, etc. The specific steps of obtaining the genomic information fields of the gene variant sites related to the variant sites include: (1) setting the query conditions, gene name: a specific gene name, for example: BRCA1, RefSNP identifier: a specific SNP identifier, for example: rs123456; (2) selecting the genomic database, NCBI dbSNP database: providing detailed information of SNP, Ensembl database: providing genomic data, including SNP and variants, other databases: UCSC Genome Browser, GnomAD, etc.; (3) constructing a query request, using the gene name and RefSNP identifier to construct a query request, for example, using the REST API interface for query; (4) extracting data, extracting the required genomic information fields of the gene variant sites from the query results, such as Gene, Cds Mutation Syntax, Amino Acid Substitution, Chromosome, Hg Position Start, Hg Position End, Ref Seq, Alt Seq, Strand, Assembly, Location, MAF value, Mechanism of Variant, etc.

[0037] In some embodiments, refer to the attached Figure 4 The scoring module 130 is used to construct a variation site evaluation system according to preset evaluation indicators, assign corresponding scoring criteria to each evaluation indicator, and calculate the score of each evaluation indicator and the comprehensive score corresponding to each variation site based on each evaluation indicator and the corresponding scoring criteria, comprehensively analyzing all target literature data; wherein a mapping relationship is formed between the preset evaluation indicators and the evaluation indicator fields.

[0038] Furthermore, the scoring module 130 includes: an evaluation system construction submodule 1310, which is used to construct a variant site evaluation system according to preset evaluation indicators, and assign corresponding scoring standards to each evaluation indicator; wherein a mapping relationship is formed between the preset evaluation indicators and the evaluation indicator fields, and the evaluation indicators include journal influence, sample size, research population, evidence level, variant site OR / HR value and number of literature reports on the variant site; a scoring calculation submodule 1320, which is used to calculate the score of each evaluation indicator and the comprehensive score corresponding to each variant site based on various evaluation indicators and corresponding scoring standards, and comprehensively integrate all target literature data.

[0039] Furthermore, based on the mapping relationship between the preset evaluation indicators and the evaluation indicator fields, a variant site evaluation system was constructed with journal influence, sample size, research population, evidence level, variant site OR / HR value and number of variant site literature reports as evaluation indicators, and each evaluation indicator was assigned a corresponding scoring standard. The specific variant site evaluation system is shown in Table 1.

[0040] Table 1 Variant site evaluation system

[0041]

[0042] Furthermore, as shown in Table 1, by integrating all target-related literature, the score of each evaluation indicator and the comprehensive score of each variant site can be calculated. In addition, for the same variant site reported by multiple literature data, the scores are calculated separately according to the different literature data, and the site screening and sorting are also involved. For the number of literature reports on the variant site, it is necessary to accumulate the number of literature reports on the variant site, score the citations obtained, and include them in the total score of the items related to the variant site to ensure the scientificity and rationality of the score.

[0043] In some embodiments, refer to the attached Figure 5 The variant site screening module 140 is used to sort the comprehensive scores of each variant site from high to low, perform site rating on the variant site according to the preset variant site rating rules, obtain the variant site rating results of each variant site, and send the scores of each evaluation indicator, the comprehensive score and the corresponding variant site rating results corresponding to each variant site to the user interaction module 150; receive and respond to the screening condition query request or the target variant site query request sent by the user interaction module 150, perform dynamic site screening on the variant site, obtain the condition screening results corresponding to the screening condition query request or the target screening results corresponding to the target variant site query request, and send the obtained condition screening results or target screening results to the user interaction module 150; and dynamically combine and name the condition screening results and / or the target screening results to create a risk site panel.

[0044] Furthermore, the variant site screening module 140 includes: a site rating submodule 1410, which is used to sort the comprehensive scores of each variant site from high to low, and to perform site rating on the variant site according to a preset variant site rating rule to obtain a variant site rating result for each variant site, and to send the score of each evaluation indicator, the comprehensive score and the corresponding variant site rating result corresponding to each variant site to the user interaction module 150; a second response submodule 1420, which is used to receive and respond to a screening condition query request or a target variant site query request sent by the user interaction module 150, perform dynamic site screening on the variant site, obtain a condition screening result corresponding to the screening condition query request or a target screening result corresponding to the target variant site query request, and send the obtained condition screening result or target screening result to the user interaction module 150; a risk site panel creation submodule 1430, which is used to dynamically combine and name the screening results according to user needs to create a risk site panel.

[0045] Furthermore, the variant sites are sorted according to the comprehensive score. The variant sites can be sorted from high to low according to the comprehensive score, and the variant sites are rated according to the preset variant site rating rules (as shown in Table 2 below); the user can select sites according to the comprehensive score from high to low according to the scoring results, or select the target variant sites of interest for dynamic combination, check the selected variant sites to form a risk site panel, name the risk site panel, and view and manage it in the Panel management submodule.

[0046] Table 2 Variant site rating rules

[0047] Variant site evaluation level Overall Rating Level I ≧30 Level II [20，30) Level III <20

[0048] In some embodiments, refer to the attached Figure 6 The user interaction module 150 is used to obtain the target risk factor keyword query request, screening condition query request and target variant site query request input by the user, send the target risk factor keyword query request to the literature retrieval module 110, send the screening condition query request and the target variant site query request to the variant site screening module 140, store, search, display and export the target literature data list, all field information of the variant site, the score of each evaluation indicator corresponding to each variant site, the comprehensive score, the corresponding variant site rating result, the condition screening result and the target screening result, and display and manage the relevant information corresponding to the created risk site Panel.

[0049] Furthermore, the user interaction module 150 includes: a Panel management submodule 1510, which is used to display, add, delete and update relevant information corresponding to the created risk site Panel, wherein the relevant information corresponding to the risk site Panel includes PanelID, name, number of sites and site information, and the site information includes the score of each evaluation indicator corresponding to the variant site, the comprehensive score, the ectopic site rating result, the target literature data related to the corresponding variant site, the evaluation indicator field related to the corresponding variant site, the key indicator field related to the corresponding variant site and the gene variant site omics information field; a site management submodule 1520, which obtains the target risk factor keyword query request, the screening condition query request and the target variant site query request input by the user, sends the target risk factor keyword query request to the literature retrieval module 110, sends the screening condition query request and the target variant site query request to the variant site screening module 140, stores, searches, displays and exports the target literature data list, all field information of the variant site, the score of each evaluation indicator corresponding to each variant site, the comprehensive score, the corresponding variant site rating result, the condition screening result and the target screening result.

[0050] Furthermore, the site management submodule also includes: a variant site submodule, which is used to obtain a target variant site query request input by a user, send the target variant site query request to the variant site screening module 140, store, search, display and export the target screening results, and store, search, display and export all field information of the variant site; a risk prediction submodule, which is used to obtain a target risk factor keyword query request and a screening condition query request input by a user, store, search, display and export a target literature data list, a score for each evaluation indicator corresponding to each variant site, a comprehensive score, the corresponding variant site rating results and the condition screening results.

[0051] Furthermore, the user interaction module 150 provides a graphical user interface (GUI), through which the user can conveniently input the target variant site, view the site information and its detailed scoring information, and also supports exporting the screening results to facilitate subsequent analysis by the user.

[0052] Furthermore, for the variant site submodule, a friendly graphical user interface is provided to allow users to input target variant sites of interest. At the same time, all variant sites obtained by the LLM large language model are stored in this submodule, and users are supported to search and view omics information based on gene, RefSNP, Cds Mutation Syntax, Hg Position Start and other gene variant site omics information fields.

[0053] Furthermore, for the risk prediction submodule, a friendly graphical user interface is provided to display detailed scoring information and related literature of the variant site, including Risk Factor, gene, RefSNP, Population, Sample size, HR / OR value, Effect Allele, Evidence level, Impact Factor, number of literature reports, Reference and other fields, as well as the score of each evaluation indicator corresponding to each variant site, the comprehensive score, and the corresponding variant site rating results; at the same time, it also supports searching and filtering based on Risk Factor, gene, RefSNP and other fields.

[0054] Embodiment 2

[0055] This embodiment provides a risk prediction variant site screening and evaluation method based on artificial intelligence on the basis of the above embodiment 1, see Figure 7 , applied to the artificial intelligence-based risk prediction variant site screening and evaluation system of Example 1, the method includes the following steps.

[0056] S1. Obtain target risk factor keyword query requests, screening condition query requests and target variant site query requests input by the user through the user interaction module, send the target risk factor keyword query requests to the literature retrieval module, and send the screening condition query requests and target variant site query requests to the variant site screening module.

[0057] S2. Receive and respond to the target risk factor keyword query request sent by the user interaction module through the document retrieval module, call the document database API through the LLM large language model, search for the target document data list related to the target risk factor keyword, and send the target document data list to the user interaction module.

[0058] S3. Extract the evaluation index fields and key indicator fields of all target document data including the variant sites from the target document data list through the evaluation index extraction module. Based on the key indicator fields, obtain the gene variant site omics information fields related to the variant sites from the omics database, and perform structured processing on the key indicator fields and the gene variant site omics information fields to obtain all field information of the variant sites stored in CSV format data, and send all field information of the variant sites to the user interaction module.

[0059] S4. A scoring module is used to construct a variation site evaluation system according to preset evaluation indicators, and each evaluation indicator is assigned a corresponding scoring standard. Based on each evaluation indicator and the corresponding scoring standard, all target literature data are integrated to calculate the score of each evaluation indicator and the comprehensive score corresponding to each variation site; a mapping relationship is formed between the preset evaluation indicators and the evaluation indicator fields.

[0060] S5. Sort the comprehensive scores of each variant site from high to low through the variant site screening module, and perform site rating on the variant site according to the preset variant site rating rules to obtain the variant site rating results of each variant site, and send the scores of each evaluation indicator, the comprehensive score and the corresponding variant site rating results corresponding to each variant site to the user interaction module; receive and respond to the screening condition query request or the target variant site query request sent by the user interaction module, perform dynamic site screening on the variant site, obtain the condition screening results corresponding to the screening condition query request or the target screening results corresponding to the target variant site query request, and send the obtained condition screening results or target screening results to the user interaction module; and dynamically combine and name the condition screening results and / or the target screening results to create a risk site panel.

[0061] S6. Store, search, display and export target literature data lists, all field information of variant sites, scores of each evaluation indicator corresponding to each variant site, comprehensive scores, corresponding variant site rating results, condition screening results and target screening results through the user interaction module, and display and manage relevant information corresponding to the created risk site panel.

[0062] Embodiment 3

[0063] This embodiment further provides an electronic device based on the above embodiment 1. Figure 8 , Figure 8 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0064] like Figure 8 As shown, the electronic device may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 to a random access memory (RAM) 303. Various programs and data required for the operation of the electronic device are also stored in the RAM 303. The processing device 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0065] Typically, the following devices may be connected to the I / O interface 305: an input device 306 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, etc., an output device 307 including, for example, a liquid crystal display (LCD), a speaker, etc., a storage device 308 including, for example, a magnetic tape, a hard disk, etc., and a communication device 309. The communication device 309 may allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Figure 8 An electronic device having various devices is shown, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 8 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0066] In particular, according to some embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from the network through the communication device 309, or installed from the storage device 308, or installed from the ROM 302. When the computer program is executed by the processing device 301, the above-mentioned functions defined in the method of some embodiments of the present disclosure are executed.

[0067] Embodiment 4

[0068] Based on the above-mentioned embodiment 1, this embodiment further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the above-mentioned method are implemented.

[0069] It should be noted that the computer-readable medium mentioned above in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In some embodiments of the present disclosure, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0070] In this embodiment, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperTextTransferProtocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an adhoc peer-to-peer network), as well as any currently known or future developed network.

[0071] The computer-readable medium may be included in the device, or may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: obtains training data, converts the training data into initial data; determines an initial rule base based on the initial data, and optimizes the parameters of the initial rule base to obtain a target rule base; calculates the rules in the target rule base according to a preset activation weight calculation formula to obtain activation weights; determines abnormal information according to the test data and the activation weights.

[0072] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0073] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0074] The units described in some embodiments of the present disclosure may be implemented by software or hardware. The units described may also be provided in a processor, for example, may be described as: a processor including a data acquisition unit, a rule determination unit, a weight calculation unit, and an anomaly determination unit. The names of these units do not, in some cases, constitute limitations on the units themselves, for example, the data acquisition unit may also be described as a "unit for acquiring training data".

[0075] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that may be used include, without limitation, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), etc.

[0076] Obviously, it should be understood by those skilled in the art that the above-mentioned various steps of the present invention can be performed in a manner different from the present invention, and the simulation method and experimental equipment include but are not limited to the above description. The above-mentioned various steps of the present invention can be performed in an order different from that here in some cases, and the steps shown or described above can be performed separately. Therefore, the present invention is not limited to any specific combination of hardware and software.

[0077] The above contents are further detailed descriptions of the present invention in combination with specific implementation methods, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the scope of protection of the present invention.

Claims

1. Risk prediction variant site screening and evaluation system based on artificial intelligence, characterized by: It includes literature retrieval module, evaluation index extraction module, scoring module, variant site screening module and user interaction module; The document retrieval module is used to receive and respond to the target risk factor keyword query request sent by the user interaction module, call the document database API through the LLM large language model, search for the target document data list related to the target risk factor keyword, and send the target document data list to the user interaction module; The evaluation index extraction module is used to extract the evaluation index fields and key indicator fields of all target document data containing variant sites from the target document data list, obtain the gene variant site omics information fields related to the variant sites from the omics database based on the key indicator fields, and perform structured processing on the key indicator fields and the gene variant site omics information fields to obtain all field information of the variant sites stored in CSV format data, and send all field information of the variant sites to the user interaction module; The scoring module is used to construct a variation site evaluation system according to preset evaluation indicators, assign corresponding scoring standards to each evaluation indicator, and calculate the score of each evaluation indicator and the comprehensive score corresponding to each variation site according to each evaluation indicator and the corresponding scoring standard, and integrate all target literature data; wherein a mapping relationship is formed between the preset evaluation indicators and the evaluation indicator fields; The variant site screening module is used to sort the comprehensive scores of each variant site from high to low, perform site rating on the variant site according to the preset variant site rating rules, obtain the variant site rating results of each variant site, and send the scores of each evaluation indicator, the comprehensive score and the corresponding variant site rating results corresponding to each variant site to the user interaction module; receive and respond to the screening condition query request or the target variant site query request sent by the user interaction module, perform dynamic site screening on the variant site, obtain the condition screening results corresponding to the screening condition query request or the target screening results corresponding to the target variant site query request, and send the obtained condition screening results or target screening results to the user interaction module; and dynamically combine and name the condition screening results and / or the target screening results to create a risk site panel; The user interaction module is used to obtain the target risk factor keyword query request, screening condition query request and target variant site query request input by the user, send the target risk factor keyword query request to the literature retrieval module, send the screening condition query request and the target variant site query request to the variant site screening module, store, search, display and export the target literature data list, all field information of the variant site, the score of each evaluation indicator corresponding to each variant site, the comprehensive score, the corresponding variant site rating result, the condition screening result and the target screening result, as well as display and manage the relevant information corresponding to the created risk site Panel.

2. The risk prediction variant site screening and evaluation system based on artificial intelligence according to claim 1, characterized in that: The document retrieval module includes: A first response submodule, configured to receive and respond to a target risk factor keyword query request sent by the user interaction module; A user input preprocessing submodule is used to perform stop word removal preprocessing and stem extraction preprocessing on the target risk factor keywords contained in the target risk factor keyword query request to obtain valid keywords; The expansion and parsing submodule is used to parse the effective keywords through the LLM large language model to obtain the parsed medical term keywords and extended synonyms related to the effective keywords; A query string generation submodule is used to generate a query string for interacting with a literature database based on parsed medical term keywords and expanded synonyms; The screening submodule is used to call the literature database through the literature database API, and screen the titles and abstracts of all literature data in the literature database based on the query string, or use a character similarity algorithm to calculate the similarity between the title or abstract of the literature data and the query string, and screen out the literature data related to the target risk factor keywords; The first formatting processing submodule is used to format the screened literature data related to the target risk factor keywords to obtain the target literature data stored in JSON format, generate a target literature data list based on the target literature data, and send the target literature data list to the user interaction module.

3. The risk prediction variant site screening and evaluation system based on artificial intelligence according to claim 2 is characterized in that: The evaluation index extraction module comprises: The document data preprocessing submodule is used to perform data cleaning and standardization on all target document data in the target document data list to obtain target document processed data; The first information extraction submodule is used to perform information extraction processing on the target literature processing data through a machine learning algorithm to extract evaluation index fields and key indicator fields; the evaluation index fields include a journal influence field, a sample size field, a research population field, an evidence level field, a variant site OR / HR value field, and a variant site literature report number field; the key indicator fields include a gene name field corresponding to the variant site, a RefSNP identifier field, a risk factor field, and an effect allele field; The second information extraction submodule is used to obtain the gene variation site omics information field related to the variation site from the omics database according to the key indicator field; the gene variation site omics information field includes a gene field, a CDS mutation syntax field, an amino acid substitution field, a chromosome field, an Hg start position field, an Hg end position field, a reference sequence field, an alternative sequence field, a chain field, a genome assembly version field, a position field, a MAF value field and a variation mechanism field; The second structured processing submodule is used to perform structured processing on the key indicator fields and the gene variation site omics information fields to obtain all field information of the variation site stored in CSV format data, and send all field information of the variation site to the user interaction module, wherein all field information of the variation site includes the key indicator fields and the gene variation site omics information fields corresponding to the variation site.

4. The risk prediction variant site screening and evaluation system based on artificial intelligence according to claim 3 is characterized in that: The scoring module includes: The evaluation system construction submodule is used to construct a variant site evaluation system according to the preset evaluation indicators, and assign corresponding scoring standards to each evaluation indicator; wherein a mapping relationship is formed between the preset evaluation indicators and the evaluation indicator fields, and the evaluation indicators include journal influence, sample size, research population, evidence level, variant site OR / HR value and number of literature reports on the variant site; The scoring calculation submodule is used to calculate the score of each evaluation indicator and the comprehensive score corresponding to each variant site based on various evaluation indicators and corresponding scoring standards, and integrate all target literature data.

5. The risk prediction variant site screening and evaluation system based on artificial intelligence according to claim 4 is characterized in that: The variant site screening module comprises: The site rating submodule is used to sort the comprehensive scores of each variant site from high to low, perform site rating on the variant site according to the preset variant site rating rules, obtain the variant site rating results of each variant site, and send the scores of each evaluation indicator, the comprehensive score and the corresponding variant site rating results of each variant site to the user interaction module; A second response submodule is used to receive and respond to the screening condition query request or the target variant site query request sent by the user interaction module, perform dynamic site screening on the variant site, obtain the condition screening result corresponding to the screening condition query request or the target screening result corresponding to the target variant site query request, and send the obtained condition screening result or target screening result to the user interaction module; The risk site panel creation submodule is used to dynamically combine and name the screening results according to user needs to create a risk site panel.

6. The risk prediction variant site screening and evaluation system based on artificial intelligence according to claim 5, characterized in that: The user interaction module comprises: The Panel management submodule is used to display, add, delete and update the relevant information corresponding to the created risk site panel, wherein the relevant information corresponding to the risk site panel includes PanelID, name, number of sites and site information, wherein the site information includes the score of each evaluation indicator corresponding to the variant site, the comprehensive score, the ectopic site rating result, the target literature data related to the corresponding variant site, the evaluation indicator field related to the corresponding variant site, the key indicator field related to the corresponding variant site and the gene variant site omics information field; The site management submodule obtains the target risk factor keyword query request, screening condition query request and target variant site query request input by the user, sends the target risk factor keyword query request to the literature retrieval module, sends the screening condition query request and the target variant site query request to the variant site screening module, stores, searches, displays and exports the target literature data list, all field information of the variant site, the score of each evaluation indicator corresponding to each variant site, the comprehensive score, the corresponding variant site rating results, the condition screening results and the target screening results.

7. The risk prediction variant site screening and evaluation system based on artificial intelligence according to claim 6 is characterized in that: The site management submodule also includes: The variant site submodule is used to obtain the target variant site query request input by the user, send the target variant site query request to the variant site screening module, store, search, display and export the target screening results, and store, search, display and export all field information of the variant site; The risk prediction submodule is used to obtain the target risk factor keyword query request and screening condition query request input by the user, store, search, display and export the target literature data list, the score of each evaluation indicator corresponding to each variant site, the comprehensive score, the corresponding variant site rating results and condition screening results.

8. A risk prediction variant site screening and evaluation method based on artificial intelligence, characterized in that: The method includes: Obtaining a target risk factor keyword query request, a screening condition query request, and a target variant site query request input by a user through a user interaction module, sending the target risk factor keyword query request to a literature retrieval module, and sending the screening condition query request and the target variant site query request to a variant site screening module; The document retrieval module receives and responds to the target risk factor keyword query request sent by the user interaction module, calls the document database API through the LLM large language model, searches for the target document data list related to the target risk factor keyword, and sends the target document data list to the user interaction module; The evaluation index extraction module extracts the evaluation index fields and key index fields of all target document data including the variant sites from the target document data list, obtains the gene variant site omics information fields related to the variant sites from the omics database based on the key index fields, and performs structured processing on the key index fields and the gene variant site omics information fields to obtain all field information of the variant sites stored in CSV format data, and sends all field information of the variant sites to the user interaction module; A scoring module is used to construct a variation site evaluation system according to preset evaluation indicators, and a corresponding scoring standard is assigned to each evaluation indicator. According to each evaluation indicator and the corresponding scoring standard, all target literature data are integrated to calculate the score of each evaluation indicator and the comprehensive score corresponding to each variation site; wherein a mapping relationship is formed between the preset evaluation indicator and the evaluation indicator field; The comprehensive scores of each variant site are sorted from high to low through the variant site screening module, and the variant sites are rated according to the preset variant site rating rules to obtain the variant site rating results of each variant site, and the scores of each evaluation indicator, the comprehensive score and the corresponding variant site rating results of each variant site are sent to the user interaction module; receiving and responding to the screening condition query request or the target variant site query request sent by the user interaction module, performing dynamic site screening on the variant sites, obtaining the condition screening results corresponding to the screening condition query request or the target screening results corresponding to the target variant site query request, and sending the obtained condition screening results or target screening results to the user interaction module; and dynamically combining and naming the condition screening results and / or the target screening results to create a risk site panel; The user interaction module is used to store, search, display and export target literature data lists, all field information of variant sites, scores of each evaluation indicator corresponding to each variant site, comprehensive scores, corresponding variant site rating results, condition screening results and target screening results, as well as display and manage relevant information corresponding to the created risk site Panel.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method as claimed in claim 8 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method as claimed in claim 8 are implemented.

Images