Analysis database about tumor somatic variation and construction method thereof

Abstract

The invention relates to an interpretation database about tumor somatic cell variation and a building method thereof. The interpretation database comprises a literature retrieval subsystem, a database management subsystem and a report generation subsystem. Based on the interpretation database provided by the invention, starting from three dimensions of medication guidance, prognosis evaluation and auxiliary diagnosis of tumor somatic cell variation, a user can quickly and accurately identify clinically meaningful variation and perform omnibearing and multi-level interpretation on the variation, and the method can be suitable for accurate treatment of tumor patients. In addition, according to the tumor somatic cell variation interpretation database provided by the invention, a powerful and abundant knowledge base system is reserved for automatically issuing reports for follow-up precise tumor medication.

Description

Technical Field

[0001] The present invention relates to the technical field of tumor database information processing, and in particular to an interpretation database for tumor somatic cell mutation and a method for constructing the database. Background Art

[0002] Genetic testing is a technology that uses modern molecular biology and molecular genetics methods to detect whether the structure of genes and their expression functions are normal. It plays an important role in the medication guidance, prognosis evaluation and auxiliary diagnosis of cancer patients. Recently, with the increasing iteration of precision medicine and targeted drugs, genetic testing is inseparable from the anti-tumor process, especially the rise of second-generation sequencing (NGS), which has greatly promoted the clinical application of genetic testing technology. Especially in clinical diagnosis and treatment related to tumors, precision medicine can detect the gene mutation spectrum of different pathological subtypes, which can assist the clinic in accurately stratifying cancer patients and provide patients with individualized treatment plans based on molecular typing, prognosis and diagnostic information.

[0003] Although next-generation sequencing has become a common detection tool for clinical diagnosis and treatment, clinicians need to further improve their ability to interpret reports. If over-interpreted or incorrectly interpreted, it will affect patients' medication guidance, prognosis assessment and auxiliary diagnosis. In order to standardize the clinical interpretation of somatic variants in cancer and support clinical decision-making, the Association for Molecular Pathology (AMP), the American Society of Clinical Oncology (ASCO) and the College of American Pathologists (CAP) jointly proposed guidelines for the interpretation of somatic variants in 2017. In order to assist clinicians in understanding the annotation and interpretation of driver gene variants, China has formulated the "Guidelines for the Interpretation of Next-Generation Sequencing Clinical Reports", and on this basis, some NGS report interpretation examples have been added. After upgrading and updating, it is called the "Consensus on the Interpretation of Tumor Next-Generation Sequencing Clinical Reports", which ultimately helps clinicians make correct interpretations and clinical decisions.

[0004] CN108897991A discloses a multi-database interactive system for head and neck tumor gene variation and drug interpretation, including: a head and neck tumor gene variation information database, a head and neck tumor drug treatment database and a head and neck tumor literature information database; each database is used to store corresponding data; among them, the head and neck tumor drug treatment database includes a head and neck tumor gene variation and drug treatment approval information database, a head and neck tumor gene variation and drug treatment expert consensus information database and a head and neck tumor gene variation and drug treatment clinical trial information database.

[0005] CN107437004A discloses a system for intelligent interpretation of personalized tumor gene detection, which includes an experimental platform, a knowledge base platform, a cloud computing platform and a terminal login platform; the experimental platform will sequence the collected biological samples on the machine, and then transmit the sequencing data to the cloud computing platform. At the same time, the user will enter the sample information and analysis requirements on the terminal login platform and then transmit the information to the cloud computing platform; the knowledge database platform is used to collect and interpret knowledge data; after the cloud computing platform performs bioinformatics analysis on the sample sequencing data and calls the data interpretation in the knowledge base platform, the generated interpretation report is transmitted to the terminal login platform, and the terminal login platform downloads the report and provides feedback.

[0006] In summary, based on the guidelines for interpretation of tumor somatic cell variants, there is currently no complete tumor somatic cell variant interpretation database that covers medication guidance, prognosis assessment, and auxiliary diagnosis. Therefore, how to establish a comprehensive, multi-level, accurate, and up-to-date tumor somatic cell variant interpretation database and automatically issue tumor precision medication reports has become one of the problems that need to be solved in this field. Summary of the invention

[0007] To solve the above technical problems, the present invention provides a database for interpreting tumor somatic cell mutations and a method for building the same. Based on the guidelines for interpreting somatic cell mutations, the database starts from three dimensions: medication guidance, prognosis assessment, and auxiliary diagnosis, to solve the problem that there is currently no complete database for interpreting tumor somatic cell mutations.

[0008] To achieve this object, the present invention adopts the following technical solutions:

[0009] In a first aspect, the present invention provides an interpretation database for tumor somatic mutations, the interpretation database comprising a literature retrieval subsystem, a database management subsystem and a report generation subsystem;

[0010] The literature retrieval subsystem is used to search and parse external databases online, and obtain the search and parsing results as evidence for gene variation, medication guidance, prognosis assessment and auxiliary diagnosis;

[0011] The database management subsystem is used to integrate evidence of gene variation, medication guidance, prognosis assessment and auxiliary diagnosis, and to perform hierarchical classification and update management. The target database is divided to store tumor gene variation, targeted therapy, chemotherapy, immunotherapy, tumor prognosis and tumor diagnosis information respectively, and interactively associates by setting key field tags, and submits and enters into the reporting system;

[0012] The report generation subsystem is used to call the target database according to the individualized data provided by the user, and automatically generate a tumor precision medication report after review and interpretation.

[0013] Preferably, the document retrieval subsystem includes:

[0014] Literature retrieval module: used to search and parse external databases online, and obtain the search and parsing results as evidence for gene variation, medication guidance, prognosis assessment and auxiliary diagnosis;

[0015] Gene mutation analysis module: used to analyze the target documents in the document retrieval module and obtain the analysis results about gene mutation;

[0016] Targeted therapy analysis module: used to analyze the target literature in the literature retrieval module and obtain analysis results on targeted therapy;

[0017] Chemical drug treatment analysis module: used to analyze the target documents in the literature search module and obtain analysis results on chemical drug treatment;

[0018] Immunotherapy analysis module: used to analyze the target documents in the literature search module and obtain analysis results about immunotherapy;

[0019] Prognosis analysis module: used to analyze the target literature in the literature retrieval module to obtain analysis results about tumor prognosis;

[0020] Diagnostic analysis module: used to analyze the target documents in the document retrieval module and obtain analysis results about tumor diagnosis.

[0021] Preferably, the database management subsystem includes:

[0022] Database architecture module: including gene variation analysis, targeted therapy and evidence level, chemotherapy and evidence level, immunotherapy and evidence level, prognosis and evidence level, and diagnosis and evidence level information;

[0023] Key field labeling module: used to assign specific labels to variant sites according to the biological functions of gene variants, and to interactively associate with medication, prognosis and diagnostic target databases to form a tumor somatic variant interpretation database with interactive matching functions;

[0024] Database classification module: used to classify the reviewed data according to the target database in the database architecture module and the somatic variation interpretation guideline standard, and store the classified target database to obtain a classified database;

[0025] Database classification module: used to divide the target database in the database classification module into clinically significant groups, and store the classified target database to obtain a classified database;

[0026] Database review module: used to send the data in the database architecture module sorted out after searching or updating by the searcher to the data administrator, so that the data administrator can review the target database;

[0027] Database submission module: used to submit each target database in the database architecture module integrated by the data administrator to the report generation subsystem;

[0028] Database update module: used to modify the information of each target database in the database architecture module and keep it dynamically updated to ensure the accuracy and timeliness of the interpretation database.

[0029] Preferably, the gene variation analysis is used to store information on the relationship between the biological function of the gene and the occurrence and development of the tumor.

[0030] Preferably, the targeted therapy and evidence level are used to store targeted drugs and drug evidence level information related to precision medicine for tumors.

[0031] Preferably, the chemotherapy treatment and evidence level are used to store chemotherapy drugs and drug evidence level information related to precision medicine for tumors.

[0032] Preferably, the immunotherapy and evidence level are used to store information on immune drugs and drug evidence levels related to precision medicine for tumors.

[0033] Preferably, the prognosis and evidence level are used to store tumor prognosis and prognosis evidence level information, and the diagnosis and evidence level are used to store tumor diagnosis and diagnosis evidence level information.

[0034] Preferably, the report generating subsystem comprises:

[0035] Sample center module: used to store and upload sample information of user-submitted items;

[0036] Experiment center module: used to store and upload user's experimental information;

[0037] Review center module: used to call the target database based on the individualized data provided by the user, and to review and interpret gene mutations;

[0038] Report center module: used to download and manage tumor precision medicine report forms with individualized data.

[0039] Preferably, the information on tumor gene mutations in the target database includes genes, transcripts, nucleotide changes, amino acid changes, mutation types, mutation functions, mutation analysis and reference data.

[0040] Preferably, the information on targeted therapy, chemotherapy and immunotherapy in the target database includes Chinese and English drug evidence, clinical trial numbers, clinical trial aliases, references, cancer names in the evidence, Chinese and English drug names, major cancer categories, evidence grading and drug efficacy data.

[0041] Preferably, the tumor prognosis information in the target database includes genes, cancer names in evidence, prognosis analysis and reference data.

[0042] Preferably, the tumor diagnosis information in the target database includes genes, cancer names in evidence, diagnostic analysis and reference data.

[0043] Preferably, the individualized data is based on the sample information and experimental information uploaded by the user and the tumor somatic mutations detected after analysis by the reporting system.

[0044] Preferably, the review and interpretation is based on the review of quality control information of the detected tumor somatic variants through the reporting system interface, while interpreting whether the detected tumor somatic variants correctly match the medication, prognosis and diagnosis information, and adjusting and optimizing the report in terms of content and format.

[0045] Preferably, the automated generation of a tumor precision medication report is accomplished by programming a tumor NGS data analysis and management system, which correlates sample information, experimental information, detected variations, quality control information, medication instructions, prognosis, and diagnosis information to automatically generate a tumor precision medication report.

[0046] In a second aspect, the present invention provides a method for building an interpretation database as described in the first aspect, the method comprising:

[0047] (1) Online search and analysis of external databases to obtain search and analysis results as evidence for gene variation, medication guidance, prognosis assessment, and auxiliary diagnosis;

[0048] (2) Integrate evidence on gene variation, medication guidance, prognosis assessment, and auxiliary diagnosis, and perform hierarchical classification and updated management;

[0049] (3) Divide the target databases to store information on tumor gene mutations, targeted therapy, chemotherapy, immunotherapy, tumor prognosis, and tumor diagnosis, and interactively associate them by setting key field labels, submitting and entering them into the reporting system;

[0050] (4) The reporting system calls the target database based on the individualized data provided by the user and automatically generates a tumor precision medication report after review and interpretation.

[0051] In a third aspect, the present invention provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implements the steps in the construction method described in the second aspect.

[0052] In a fourth aspect, the present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program or instructions, and when the computer program or instructions are executed by the processor, the steps in the construction method described in the second aspect are implemented.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] The present invention provides an interpretation database for tumor somatic variation and a method for building it according to the guidelines for interpretation of tumor somatic variation. Starting from the three dimensions of medication guidance, prognosis assessment and auxiliary diagnosis of tumor somatic variation, the clinical significance of tumor somatic variation is evaluated in an all-round, multi-level, accurate and up-to-date manner to the greatest extent; and further combined with the literature retrieval subsystem, database management subsystem and report generation subsystem to form a set of tumor somatic variation interpretation databases with interactive matching, which are suitable for the precision treatment of tumor patients. In addition, the tumor somatic variation interpretation database provided by the present invention also reserves a powerful and rich knowledge base system for the subsequent automated issuance of reports for precision tumor medication. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A schematic diagram of the structure of the database for interpretation;

[0056] Figure 2 This is the distribution map of the clinical significance of tumor somatic mutations;

[0057] Figure 3 A flowchart for explaining the database construction method. DETAILED DESCRIPTION

[0058] To further illustrate the technical means and effects of the present invention, the present invention is further described below in conjunction with the embodiments and drawings. It should be understood that the specific implementation methods described herein are only used to explain the present invention, rather than to limit the present invention.

[0059] A schematic diagram of the structure of a database for interpreting tumor somatic cell mutations provided by the present invention is as follows: Figure 1As shown, it includes a literature retrieval subsystem, a database management subsystem, and a report generation subsystem. The literature retrieval subsystem is used to search and parse external databases online, and obtain the retrieval and analysis results as evidence for gene variation, medication guidance, prognosis evaluation, and auxiliary diagnosis; the database management subsystem is used to integrate, classify, and update the retrieval and analysis results in the literature retrieval subsystem, and divide multiple target databases to store information such as tumor gene variation, targeted therapy, chemotherapy, immunotherapy, tumor prognosis, and tumor diagnosis, and interactively associate them by setting key field labels, submitting and entering them into the report system; the report generation subsystem is used to call the target database based on the individualized data provided by the user, and automatically generate a tumor precision medication report after review and interpretation.

[0060] In practical applications, the literature retrieval subsystem includes a literature retrieval module; the literature retrieval module is used to search and parse external databases online, and obtain the search and parsing results as evidence for gene variation, medication guidance, prognosis evaluation and auxiliary diagnosis. Among them, online retrieval and parsing of external databases refers to manually searching external databases such as Oncokb database (https: / / www.oncokb.org / ), CKB database (https: / / ckb.jax.org / ), FDA (https: / / www.fda.gov / ), NMPA (http: / / www.nmpa.gov.cn / ), clinical trials (https: / / clinicaltrials.gov / ), PharmGKB database (https: / / www.pharmgkb.org / ), Google (https: / / scholar.google.hk / ), PubMed (https: / / pubmed.ncbi.nlm.nih.gov / ) and other platform resources based on gene variation information, and obtaining target information such as tumor gene variation, drugs, prognosis and diagnosis after parsing, and temporarily storing this target information in a temporary data table for subsequent detailed analysis.

[0061] In actual applications, the literature retrieval subsystem also includes a gene variation analysis module; the gene variation analysis module is used to analyze the target information in the literature retrieval module to obtain analysis results about gene variation. Specifically, the biological function information of gene variation is queried through the Oncokb database, CKB database, Google / PubMed literature search, etc. The biological functions of somatic cell variation are divided into four categories: ① gain-of-function mutations (Gain of function, GOF), loss-of-function mutations (Loss of function, LOF), unknown functions (unknown) and no effects (no effect). Synchronously extract information such as genes, transcripts, nucleotide changes (c.), amino acid changes (p.), variation types, variation functions, variation analysis, references, etc. into a list for subsequent calls to the database architecture module. The specific carcinogenicity judgment of tumor somatic variation is as follows: Figure 2 shown.

[0062] In actual applications, the literature retrieval subsystem also includes a targeted therapy analysis module; the targeted therapy analysis module is used to analyze the target literature in the literature retrieval module to obtain analysis results about targeted therapy. Specifically, it enters the FDA / NMPA official website, NCCN / CSCO guidelines, clinical trials ClinicalTrials, Google / Pubmed and other platforms to query drug information at gene mutation sites, and simultaneously extracts drug evidence (in Chinese and English), clinical trial number (NCT), clinical trial alias, references, cancer names in evidence, drug names (in Chinese and English), major cancer categories, evidence classification, drug efficacy and other information into a list for subsequent database architecture module calls.

[0063] In actual applications, the literature retrieval subsystem also includes a chemical drug treatment analysis module; the chemical drug treatment analysis module is used to analyze the target literature in the literature retrieval module to obtain analysis results about chemical drug treatment. Specifically, it enters the FDA / NMPA official website, NCCN / CSCO guidelines, PharmGKB database and other platforms to query chemotherapy drug information, such as searching the PharmGKB database based on the RS number of known genes, and simultaneously extracting drug names (in Chinese and English), drug toxicity / sensitivity and other information into the list for subsequent database architecture module calls.

[0064] In actual applications, the literature retrieval subsystem also includes an immunotherapy analysis module; the immunotherapy analysis module is used to parse the target literature in the literature retrieval module to obtain analysis results about immunotherapy. Specifically, it enters the FDA / NMPA official website, NCCN / CSCO guidelines, clinical trials ClinicalTrials, Google / Pubmed and other platforms to query drug information at gene mutation sites, and simultaneously extracts drug evidence (Chinese and English), clinical trial number (NCT), clinical trial alias, references, cancer names in evidence, drug names (Chinese and English), major cancer categories, evidence classification, drug efficacy and other information into a list for subsequent database architecture module calls.

[0065] In actual applications, the literature retrieval subsystem also includes a prognosis analysis module; the prognosis analysis module is used to analyze the target literature in the literature retrieval module to obtain analysis results about tumor prognosis. Specifically, it enters the NCCN / CSCO guidelines and Google / Pubmed platforms to query the prognosis information of gene mutation sites, and simultaneously extracts genes, cancer names in evidence, prognosis analysis, references and other information into a list for subsequent database architecture module calls.

[0066] In actual applications, the literature retrieval subsystem also includes a diagnostic analysis module; the diagnostic analysis module is used to parse the target literature in the literature retrieval module to obtain analytical results on tumor diagnosis. Specifically, it enters the NCCN / CSCO guidelines and Google / Pubmed platforms to query the diagnostic information of gene mutation sites, and simultaneously extracts genes, cancer names in evidence, diagnostic analysis, references and other information into a list for subsequent database architecture module calls.

[0067] In actual applications, the database management subsystem includes a database architecture module; the database architecture module is used to store the analysis content obtained from each target database in the literature retrieval subsystem after review, and on this basis, it is graded and classified based on the somatic variation interpretation guideline standards to obtain gene variation analysis, targeted therapy and evidence level, chemotherapy and evidence level, immunotherapy and evidence level, prognosis and evidence level, diagnosis and evidence level, among which each target database module exists independently.

[0068] In actual applications, the database management subsystem also includes a database classification module; the database classification module is used to classify the reviewed data based on the somatic variation interpretation guide standards according to the various target databases in the database architecture module, and store the classified target databases to obtain a classified database; it includes targeted therapy and evidence level, chemical drug therapy and evidence level, immunotherapy and evidence level, prognosis and evidence level, diagnosis and evidence level.

[0069] Among them, targeted and immunotherapy and their evidence levels include A, B, C and D; Grade A indicates that the drug sensitivity evidence is approved by the FDA / NMPA or recommended by professional guidelines; Grade B indicates that the drug sensitivity evidence is confirmed by clinical studies with sufficient statistical power and has obtained expert consensus in this field; Grade C indicates that the drug sensitivity evidence is Level A evidence in other cancer types (cross-indication use), supported by multiple small clinical trials that have been carried out, and has been used as an inclusion criterion for clinical trials; Grade D indicates that the drug sensitivity evidence is supported by a single case report or preclinical evidence. Chemotherapy and its evidence levels include level 1A, level 1B, level 2A, level 2B, level 3 and level 4; 1A means that there is a lot of clinical data support, and the clinical guidelines or FDA explain the specific mutant alleles in detail, focusing on the analysis of the impact of single nucleotide mutations on genotype and phenotype, which is the highest standard for clinical application; 1B means that there is high-quality evidence support, but the clinical guidelines or FDA do not explain the specific mutant alleles, and the evidence level is lower than 1A; 2A and 2B are medium evidence levels. 2B means that there is medium evidence support in multiple repeated studies, but some of these studies are not statistically significant or have a small effect size. 2A means that on the basis of meeting 2B, it must be defined as a very important drug gene in PharmGKB, and this gene may have functional significance; level 3 is a low evidence level, that is, it is shown to be significant in a single study (no reproducibility), or there are multiple studies, but there is a lack of clear evidence to show that it is related (no significant difference); level 4 is the level of evidence against, indicating that the existing evidence does not support the relevance of gene drug therapy and further verification is needed. Prognosis and diagnosis and their evidence levels include A, B, C, and D. A represents diagnostic / prognostic factors for specific tumors as defined in professional guidelines; B represents diagnostic / prognostic values ​​confirmed by clinical studies with sufficient statistical power; C represents diagnostic / prognostic values ​​supported by multiple small studies; and D represents small studies or case reports suggesting auxiliary diagnostic / prognostic values ​​(independently or in combination with other markers). The evidence levels and sources for specific tumor somatic mutations regarding medication, prognosis, and diagnosis are as follows: Figure 2 shown.

[0070] In practical applications, the database management subsystem also includes a database classification module; the database classification module is used to divide the target database in the database grading module according to clinical significance, and store the classified target database to obtain a classified database; tumor somatic mutations are divided into four levels according to their clinical significance: Class I mutations (important clinical significance): including Class A and Class B evidence (expert consensus belongs to Class B evidence); Class II mutations (potential clinical significance): including Class C and Class D evidence (for mutations of unknown functional type, if there are cases of drug sensitivity, co-mutation detection or drug resistance, the clinical significance is determined to be Class II); Class III mutations (clinical significance is unknown): there is no high incidence in both population databases and tumor-related databases, and there is no confirmed literature evidence related to tumors; Class IV mutations (benign / likely benign): the mutation frequency is high in population databases, and there is no literature evidence related to tumors.

[0071] According to the interpretation standards for tumor somatic variation, if the target site to be queried falls within the scope of FDA / NMPA approval and NCCN / CSCO recommendation, level A evidence is given and the clinical significance is judged to be a Class I variation; if it does not fall within the approved / recommended scope, level B, level C, and level D evidence are given respectively, and the clinical significance is judged to be a Class I variation (level B) and Class II variation (level C and level D); if it does not fall within the approved / recommended scope and its function is unknown, its clinical significance is judged to be a Class III variation; if it does not fall within the approved / recommended scope, its function has no effect and there are no case reports (drug sensitivity, co-mutation detection, or drug resistance), its clinical significance is judged to be a Class IV variation. The specific clinical significance of tumor somatic variation regarding medication, prognosis, and diagnostic evidence is divided as follows: Figure 2 shown.

[0072] In actual applications, the database management subsystem also includes: key field label module, database review module, database submission module, and database update module. Among them, the key field label is a specific label assigned to the mutation site based on the biological function of the gene mutation, and interactively associates with target databases such as medication, prognosis and diagnosis to form a set of tumor somatic mutation interpretation databases with interactive matching. The database review module is used to send the data in the database architecture module sorted after the searcher retrieves / updates to the data administrator to realize the data administrator's review of the target database. The database submission module is used to submit each target database in the database architecture module integrated by the data administrator to the report generation subsystem. The database update module is used to modify the information of each target database in the database architecture module and keep it dynamically updated to ensure the accuracy and timeliness of the interpretation database.

[0073] Example 1

[0074] In order to obtain interpretation data on the biological function, targeted therapeutic drugs, prognosis or diagnostic information of EGFR L858R gene mutation in lung cancer patients, the process of using the interpretation database provided by the present invention is as follows:

[0075] (1) In the Gene Variation Analysis module, filter EGFR in the "Gene" list, click the "Amino Acid Change (p.)" list to filter L858R, and find the "Variation Analysis" list to obtain the biological function information about the EGFR L858R gene variation, and simultaneously obtain the specific label corresponding to the EGFR L858R gene variation, which is "EGFR_MGT". The gene variation analysis is shown below.

[0076] For example, the EGFR exon 21 L858R mutation occurs in the EGFR tyrosine kinase domain. Cell line experiments have shown that this mutation has activating, transforming and sensitizing effects on EGFR tyrosine kinase inhibitors (TKIs) (PMID: 15118073, 15118125). Preclinical models of the L858R mutation are highly sensitive to first-generation, second-generation and third-generation inhibitors as well as ex20in activity inhibitors (PMID: 34526717).

[0077] (2) Lung cancer was screened in the "Major cancer types" and "Cancer types in evidence" lists of the targeted therapy and evidence level module. Based on the specific labels obtained, the targeted therapy drugs corresponding to "EGFR_MGT" (such as sensitive A-level osimertinib) were found. The drug efficacy information about the EGFR L858R gene mutation was obtained in the "Drug evidence (Chinese and English)" list, and the drug evidence level and references were obtained simultaneously. The drug efficacy information is shown below.

[0078] For example, the FDA and NMPA have approved osimertinib for the first-line treatment of adult patients with metastatic NSCLC carrying EGFR Exon 19del or Exon 21L858R mutations, and for postoperative adjuvant treatment of patients with surgically resectable NSCLC with EGFR exon 19 deletion or exon 21 (L858R) mutations.

[0079] For example, the "NCCN Clinical Practice Guidelines: Non-Small Cell Lung Cancer (2024.V9)" suggests that high-risk non-small cell lung cancer (NSCLC) patients with IB-IIA tumors and negative margins after tumor resection with EGFR exon 19 deletion or exon 21L858R mutation should receive chemotherapy and adjuvant osimertinib; patients with II tumors and negative margins after tumor resection with EGFR exon 19 deletion or exon 21L858R mutation should receive chemotherapy and adjuvant osimertinib; if EGFR exon 19 deletion or exon 21L858R mutation is detected before first-line treatment, osimertinib (preferred) is recommended for the first-line treatment of patients with advanced / metastatic NSCLC; if EGFR exon 19 deletion or exon 21L858R mutation is detected during first-line systemic treatment, the originally planned systemic treatment, including maintenance therapy, should be completed or interrupted, and osimertinib (preferred) is recommended. If there is asymptomatic disease progression or disease progression in the brain or limited systemic metastasis after osimertinib treatment, the guidelines recommend continuing osimertinib treatment. For patients with non-small cell lung cancer with EGFR exon 19 deletion or exon 21 L858R mutation, if the disease progresses during first-line treatment with erlotinib (± ramucirumab or bevacizumab) or afatinib or gefitinib or dacomitinib, T790M testing is recommended. If T790M+, asymptomatic or with brain lesions, osimertinib follow-up treatment can be considered. Multiple lesions that have not been given osimertinib can also be selected for osimertinib follow-up treatment. The guidelines suggest: For progressive CNS (central nervous system) disease or leptomeningeal disease, osimertinib can be considered (regardless of T790M status).

[0080] For example, the "CSCO Guidelines for the Diagnosis and Treatment of Non-Small Cell Lung Cancer (2024)" recommends: For patients with stage IIA / IIB non-small cell lung cancer (NSCLC) who are tested positive for EGFR sensitive mutations after radical surgery, osimertinib should be used after surgery (after adjuvant chemotherapy) (Grade I recommendation); for patients with stage IIIA / B radical surgery: if postoperative pathological detection shows EGFR sensitive mutations, osimertinib should be used after surgery (after adjuvant chemotherapy) (Grade I recommendation); Osimertinib (Grade I recommendation) is recommended as the first-line treatment for patients with advanced NSCLC carrying EGFR sensitive mutations.

[0081] (3) In the "Cancer Types in Evidence" list of the Tumor Prognosis and Evidence Level module, select lung cancer, and find the prognostic analysis corresponding to "EGFR_MGT" based on the specific label obtained. You can obtain information such as the prognostic evaluation and references of the EGFR L858R gene mutation. The prognostic information is as follows:

[0082] For example, the study by Takamochi et al. showed that among 939 surgically resected lung adenocarcinoma patients (79% of whom were in stage I and 21% in stage II-IV), univariate analysis showed that patients with EGFR mutations had significantly better overall survival (OS) and recurrence-free survival (RFS), and in multivariate analysis, EGFR mutations were proven to be an independent good prognostic factor for OS (PMID: 28826599). Regarding the difference in the impact of different EGFR mutation subtypes on prognosis, some previous studies have given some preliminary conclusions, but these conclusions are not consistent. In the study by Takamochi et al., among 418 surgically resected EGFR mutation lung adenocarcinoma patients, 222 (53%) carried EGFR L858R and 161 (39%) carried EGFR 19del. The results showed that there was no difference in RFS and OS among different EGFR mutation subtypes (PMID: 28826599). In a meta-analysis of 4,872 patients, researchers analyzed the data of 1,020 patients with surgically resected stage I-III EGFR-mutant NSCLC who had not received neoadjuvant or adjuvant EGFR-TKI therapy. The results showed that the EGFR 19del subgroup had a worse disease-free survival (DFS) than the EGFR L858R subgroup, but it did not reach statistical significance (PMID: 31106123). Another nationwide retrospective study in Japan included 2,410 patients with surgically resected EGFR mutations, of whom 1,170 (48.5%) carried EGFR L858R and 983 (40.8%) carried EGFR19 del. The results showed that, in both univariate and multivariate analysis, EGFR 19del patients had significantly worse RFS than EGFR L858R patients, EGFR 19del had a more aggressive phenotype, and the prognosis of early lung cancer patients was worse than L858R (PMID: 32615091). Based on the above clinical studies, the "Expert Consensus on Prediction of Postoperative Recurrence of Non-Small Cell Lung Cancer Based on Molecular Markers" suggests that among the common EGFR mutations, patients with EGFR 19del have a higher risk of postoperative recurrence (grade 2A) than those with EGFR L858R mutations.

[0083] (4) In the "Cancer Types in Evidence" list of the Tumor Diagnosis and Evidence Level module, select lung cancer, and find the diagnostic analysis corresponding to "EGFR_MGT" based on the specific tags obtained. You can obtain information such as auxiliary diagnosis and references about EGFR L858R gene mutation. The diagnostic information is shown below.

[0084] For example, EGFR mutation is the most common mutation type in patients with non-small cell lung cancer (NSCLC). Studies have shown that about 50% of Asian NSCLC patients have EGFR mutations, and their EGFR mutation rate is higher than that of Western NSCLC patients (PMID: 19692684). The NCCN Clinical Practice Guidelines: Non-Small Cell Lung Cancer (2024.V9) points out that the two most common EGFR mutations in NSCLC are 45% exon 19 deletions (conservative deletions of the LREA sequence) and 40% exon 21 point mutations (L858R). These two mutations account for about 85%-90% of all EGFR mutations in NSCLC. Both can activate the tyrosine kinase domain and are associated with sensitivity to small molecule EGFR tyrosine kinase inhibitors (TKIs). The guidelines suggest that the predictive role of the EGFR L858R mutation is clear, and patients carrying this mutation respond significantly better to EGFR TKIs such as afatinib, dacomitinib, erlotinib, gefitinib, or osimertinib. "ESMO: Expert Consensus on the Treatment of EGFR-Mutated Non-Small Cell Lung Cancer (2022 Edition)" recommends that it is best to use second-generation sequencing to perform extensive testing of exons 18-21 of the EGFR tyrosine kinase region in order to discover all mutations with clear or potential clinical utility (evidence level: I, recommendation level: A).

[0085] Example 2

[0086] In order to obtain interpretation data on immunotherapy drugs and prognostic information for colorectal cancer patients with high microsatellite instability (MSI-H), the process of using the interpretation database provided by the present invention is as follows:

[0087] (1) In the "Gene" list of the gene variation analysis module, MSI is screened, and the variation analysis of the immune biomarker MSI-H is found. At the same time, the specific label of high microsatellite instability is known as "MSI_H". The gene variation analysis is shown below.

[0088] For example, microsatellite instability (MSI) refers to the phenomenon of an increase or decrease in the base length and / or number of repetitions of short repeat sequences in DNA sequences. Functional mutations in genes related to the mismatch repair (MMR) system in tumor cells or hypermethylation of the MLH1 promoter region lead to abnormal function of the MMR protein, which is unable to repair errors that occur during DNA replication and ultimately leads to microsatellite instability. Studies have shown that MSI-H is found in a variety of solid tumors, with an overall incidence of 3.8%, including endometrial cancer (31.4%), colorectal cancer (19.72%) and gastric cancer (19.09%) (PMID: 29850653).

[0089] (2) In the "Major cancer types" and "Cancer types in evidence" lists of the immunotherapy and evidence level module, colorectal cancer was screened, and based on the immune biomarker MSI, the immunotherapy drugs corresponding to "MSI_H" (such as sensitive A-level pembrolizumab) were found. The drug efficacy information about MSI_H was obtained in the "Drug evidence (Chinese and English)" list, and the drug evidence level and references were obtained simultaneously. Its immune information is shown below.

[0090] For example, the FDA has accelerated the approval of pembrolizumab (K drug) for the treatment of adult and pediatric patients with unresectable or metastatic MSI-H / dMMR solid tumors who have progressed after previous treatment and have no satisfactory alternative treatment options. The FDA has accelerated the approval of pembrolizumab for the treatment of patients with colorectal cancer who have progressed after fluoropyrimidine, oxaliplatin, and irinotecan treatment; the FDA and NMPA have approved pembrolizumab as a single agent for the first-line treatment of patients with wild-type KRAS, NRAS, and BRAF genes, unresectable or metastatic high microsatellite instability (MSI-H) or mismatch repair gene deficiency (dMMR) colorectal cancer (CRC).

[0091] For example, the NCCN Clinical Practice Guidelines: Colon Cancer (2024.V5) recommends immune checkpoint inhibitors (including nivolumab ± ipilimumab, pembrolizumab) for neoadjuvant treatment of patients with dMMR / MSI-H (non-metastatic) clinical T4b colon cancer (preferred); for patients with dMMR / MSI-H and resectable synchronous simple liver and / or lung metastatic colon cancer, if immunotherapy has not been performed before, immune checkpoint inhibitors (including nivolumab ± ipilimumab, pembrolizumab, dostarlimab-gxly) (preferred) can be considered, followed by simultaneous or staged colectomy and resection of metastatic disease (preferred) and / or local therapy. "NCCN Clinical Practice Guidelines: Rectal Cancer (2024.V4)" recommends nivolumab or pembrolizumab or dostarlimab-gxly as neoadjuvant treatment / definitive immunotherapy (preferred) for patients with dMMR / MSI-H and (T3, N any) or (T1-2, N1-2) or (T4, N any) or locally unresectable or medically inoperable rectal cancer; if immunotherapy has not been performed before, immune checkpoint inhibitors (including nivolumab ± ipilimumab, pembrolizumab, dostarlimab-gxly) (preferred) are recommended as neoadjuvant treatment options for patients with dMMR / MSI-H and resectable synchronous simple liver and / or lung metastases to the rectal cancer. "NCCN Clinical Practice Guidelines: Colon Cancer (2024.V5) / Rectal Cancer (2024.V4)" For patients with dMMR / MSI-H and resectable metachronous metastases, if they have not received immunotherapy before, immune checkpoint inhibitors (including nivolumab ± ipilimumab, pembrolizumab, dostarlimab-gxly) can be considered for initial treatment, followed by observation, resection or local treatment; for patients with dMMR / MSI-H advanced or metastatic colon cancer who have not received immunotherapy before, immune checkpoint inhibitors (including nivolumab ± ipilimumab, pembrolizumab, dostarlimab-gxly) are recommended for systemic treatment, among which (nivolumab + ipilimumab) is recommended as a 2B category. Due to toxicity issues, intensive treatment is not recommended. The guidelines recommend that patients undergoing immunotherapy should be closely followed up for 10 weeks to evaluate the response. The guidelines suggest that if the disease is in remission, it is possible to consider discontinuing checkpoint immunosuppressants after 2 years of treatment. The guidelines suggest that patients with dMMR / MSI-H or POLE / POLD1 mutations who are not suitable for immunotherapy should be treated according to the recommended treatment methods for pMMR / MSS disease.

[0092] For example, the "CSCO Colorectal Cancer Diagnosis and Treatment Guidelines (2023)" and the "CSCO Immune Checkpoint Inhibitor Clinical Application Guidelines (2023)" recommend pembrolizumab (Grade I recommendation) as the first-line treatment for the palliative treatment group of patients with metastatic colorectal cancer carrying MSI-H / dMMR; immune checkpoint inhibitors (pembrolizumab, envobalizumab, slulizumab, tislelizumab) (Grade II recommendation) are recommended as palliative second-line treatment (immune checkpoint inhibitors were not used in the first line) and third-line (immune checkpoint inhibitors were not used in the first / second line) treatment options for patients with metastatic colorectal cancer carrying MSI-H / dMMR.

[0093] (3) In the "Cancer Types in Evidence" list of the Tumor Prognosis and Evidence Level module, select colorectal cancer, and find the prognostic analysis corresponding to "MSI_H" based on the specific label obtained. You can obtain various information such as the prognostic analysis and references about MSI_H. The prognostic information is shown below.

[0094] For example, the NCCN Clinical Practice Guidelines: Colon Cancer (2024.V5) suggests that data from the PETACC-3 trial show that MSI-H is more common in patients with stage II tumors than in patients with stage III (22% vs 12%, P<0.0001) (PMID: 20008640). In another large study, MSI-H was only 3.5% in patients with stage IV tumors (PMID: 19165197). These results suggest that tumors carrying MSI-H (dMMR) are less likely to metastasize. In fact, a large body of evidence suggests that in patients with stage II tumors, MSI-H (dMMR) tumor status is a prognostic marker for more favorable outcomes (PMID: 25361982, 12867608, 20498393). In contrast, the beneficial effects of MSI-H (dMMR) appear to be more limited in stage III colon cancer and may vary by primary tumor location (PMID: 25361982, 24019539). Some of these same studies also suggest that MSI-H (dMMR) tumor status may be a predictive marker for reduced benefit from adjuvant fluoropyrimidine therapy alone in patients with stage II tumors and may have deleterious effects (PMID: 12867608, 20498393, 26271397). In contrast to the findings of Sargent et al. (PMID: 20498393), a study of 1913 patients with stage II CRC by the QUASAR study showed that although dMMR was a prognostic factor (recurrence rate: dMMR 11% vs pMMR 26%), it did not predict benefit or deleterious effects of chemotherapy (PMID: 21383284). Studies of patients in the CALGB9581 and 89803 trials reached similar conclusions that MMR status is a prognostic indicator but does not predict the benefit or harm of adjuvant therapy (irinotecan combined with 5-FU / LV (IFL regimen)) in patients with stage II cancer (PMID: 21747089). Guidelines suggest that patients with stage II MSI-H tumors may have a good prognosis and not benefit from adjuvant 5-FU therapy, and that patients with low-risk stage II MSI-H tumors should not receive adjuvant therapy. It should be noted that poorly differentiated histology is not considered a high-risk feature for patients with stage II disease whose tumors are MSI-H.

[0095] Example 3

[0096] In order to obtain interpretation data of biological functions, targeted therapeutic drugs, prognosis or diagnostic information of KRAS G12D gene mutations in thyroid cancer patients, the process of using the interpretation database provided by the present invention is as follows:

[0097] (1) In the Gene Variation Analysis module, filter KRAS in the "Gene" list, click the "Amino Acid Change (p.)" list to filter G12D, and find the "Variation Analysis" list to obtain the biological function information about the KRAS G12D gene variation, and simultaneously obtain the specific label corresponding to the KRAS G12D gene variation as "KRAS_G12D". The gene variation analysis is shown below.

[0098] For example, the KRAS G12D mutation is located in the P loop of the catalytic G domain of the protein. This mutation has been found in lung cancer, colorectal cancer, pancreatic cancer, and ovarian cancer (PMID: 28572459). Compared with the wild type, the expression of this mutation in cell lines and mouse models showed that the mutation was activated by increasing downstream pathway activation, colony formation, and multi-lineage in vivo tumor development (PMID: 20570890, 20147967, 11751631, 15093544, 19296721, 17349581, 32792368).

[0099] (2) Lung cancer was screened in the "Major cancer types" and "Cancer types in evidence" lists of the targeted therapy and evidence level module. Based on the specific labels obtained, the targeted therapy drugs corresponding to "KRAS_G12D" (such as sensitive D-level sorafenib) were found. The drug efficacy information about KRAS_G12D gene mutations was obtained in the "Drug evidence (Chinese and English)" list, and the drug evidence level and references were obtained simultaneously. The drug efficacy information is shown below.

[0100] For example, in a phase II clinical trial, 57 patients with advanced non-small cell lung cancer (NSCLC) who carried KRAS mutations (G12C, n=31; G12D, n=8; G12V, n=8; G13C, n=1; G13Y, n=1; Q61L, n=2; G12A, n=3; G12F, n=1; G13A, n=1; G12S, n=1) and had previously received at least one platinum-containing treatment were treated with sorafenib. At week 6, 5 patients achieved partial remission, 25 patients had stable disease, and 27 patients had progressive disease, with a disease control rate of 52.6%. The median progression-free survival (PFS) was 2.3 months, and the median overall survival (OS) was 5.3 months. Studies have shown that sorafenib has relevant clinical activity in NSCLC patients carrying KRAS mutations. (PMID: 23224737)

[0101] (3) In the "Cancer Types in Evidence" list of the Tumor Prognosis and Evidence Level module, select lung cancer, and find the prognostic analysis corresponding to "KRAS_G12D" based on the specific label obtained. You can obtain information such as prognostic evaluation and references about KRAS_G12D gene mutations. The prognostic information is shown below.

[0102] As pointed out in the "Guangdong Expert Consensus on Genetic Testing and Clinical Application of Thyroid Cancer (2020 Edition)", it has been found that RAS and TERT mutations have synergistic effects, and double gene mutations are associated with FTC recurrence and increased mortality; BRAF / RAS combined with PIK3CA, TP53, EIF1A and other mutations indicate high tumor invasiveness and poor prognosis, which is closely related to iodine resistance. The "Bethesda Thyroid Cytology Classification Diagnostic System (2023 Edition)" points out that tumors can be stratified into low, medium, and high molecular risk groups (MRG). Low MRG manifests as a single RAS mutation or RAS-like variation. Medium MRG includes BRAFV600E mutations, other BRAF-like variations, or copy number changes. The high MRG feature is that one of the above-mentioned driver changes co-occurs with gene mutations [such as TERT, TP53, AKT1 and (or) PIK3CA]; this stratification helps to identify thyroid cancer subgroups with poor prognosis.

[0103] (4) In the "Cancer Types in Evidence" list of the Tumor Diagnosis and Evidence Level module, select lung cancer, and find the prognostic analysis corresponding to "KRAS_G12D" based on the specific label obtained. You can obtain various information such as auxiliary diagnosis and references about KRAS_G12D gene mutations. The diagnostic information is shown below.

[0104] For example, the "Guidelines for the Diagnosis and Treatment of Thyroid Cancer (2022 Edition)" points out that RAS point mutations are common gene mutations in follicular thyroid cancer (FTC). The "Guangdong Expert Consensus on Genetic Testing and Clinical Application of Thyroid Cancer (2020 Edition)" points out that the mutation rate of RAS in differentiated thyroid cancer is second only to BRAF, which is 40%-50% in FTC and 10%-20% in PTC. The mutation rate is highest in follicular subtype papillary thyroid cancer (FVPTC), and the mutation rate in non-invasive follicular thyroid tumors (NIFTP) with papillary cell nuclear features is 20%-40%, and 59% of cases with RAS mutations are NIFTP. RAS mutations have also been found in benign thyroid nodules, so there are limitations in the use of RAS for the definitive diagnosis of thyroid nodules. The "Guidelines for the Diagnosis and Treatment of Thyroid Nodules and Differentiated Thyroid Cancer (Second Edition)" points out that RAS gene mutations are common in FTC (50%); RAS gene mutations are found in some oncocytic carcinomas (but the probability is lower than in FTC); most differentiated high-grade thyroid cancers are driven by BRAF V600E mutations, but some are also driven by RAS gene mutations. The "Bethesda Thyroid Cytology Classification Diagnostic System (2023 Edition)" points out that mutations in the proto-oncogene RET are the cause of most MTCs, and a small number of sporadic MTCs are caused by HRAS, KRAS, and NRAS mutations. In FTC, the most common mutations are in the RAS gene family. Encapsulated invasive papillary carcinoma follicular variant (IEFVPTC) is a RAS-like tumor that is closer to follicular thyroid carcinoma (FTC) than PTC.

[0105] Example 4

[0106] In order to obtain the interpretation data of the drug sensitivity or toxic side effect information of the patient's chemotherapy gene UGT1A1 rs3064744 site, the process of using the interpretation database provided by the present invention is as follows:

[0107] In the "Gene" list of the chemical drug treatment and evidence level module, filter UGT1A1 and simultaneously filter the rs3064744 locus. According to different typing results (TA7 / TA7, TA6 / TA7, TA6 / TA6), you can get the drug sensitivity and side effect information, and the drug evidence level. The typing results are derived from the PharmGKB database, and different chemotherapy genes have different typing results. The chemotherapy information is shown below.

[0108] For example, when screening the rs3064744 locus of the UGT1A1 gene, when the test result is TA6 / TA6, the toxicity of FOLFIRI is that the risk of toxicity may be low (evidence level 1A); the toxicity of FOLFIRINOX is that the risk of toxicity may be low (evidence level 3); the drug sensitivity of irinotecan is that the dose of irinotecan may need to be increased (evidence level 1A). When the test result is TA6 / TA7, the toxicity of FOLFIRI is that the risk of toxicity may be high (evidence level 1A); the toxicity of FOLFIRINOX is that the risk of toxicity may be high (evidence level 3); the drug sensitivity of irinotecan is that the dose of irinotecan may need to be reduced (evidence level 1A). When the test result is TA7 / TA7, the side effects of FOLFIRI are that the risk of side effects may be high (evidence level is 1A); the side effects of FOLFIRINOX are that the risk of side effects may be high (evidence level is 3); the drug sensitivity of irinotecan is that the dose of irinotecan may need to be reduced (evidence level is 1A).

[0109] A schematic diagram of a method for building a tumor somatic cell variation interpretation database provided by the present invention is as follows: Figure 3 As shown. The construction method of the interpretation database includes 4 modules: ① Online retrieval and analysis of external databases, and obtaining retrieval and analysis results as evidence for gene variation, medication guidance, prognosis evaluation and auxiliary diagnosis. ② Integrate the evidence of gene variation, medication guidance, prognosis evaluation and auxiliary diagnosis, and classify and update them. ③ Divide multiple target databases to store information such as tumor gene variation, targeted therapy, chemotherapy, immunotherapy, tumor prognosis and tumor diagnosis, and interactively associate them by setting key field labels, submitting and entering them into the reporting system. ④ The reporting system calls the target database based on the individualized data provided by the user, and automatically generates a tumor precision medication report after review and interpretation.

[0110] In practical applications, since the construction method is applied to the interpretation database of tumor somatic cell mutations, each step corresponds to the literature review subsystem, the database management subsystem, and the report generation subsystem. The first to third modules have been described in detail in the above embodiments and will not be repeated here. The following mainly describes the embodiment of the report generation subsystem in detail.

[0111] In actual applications, the report generation subsystem includes a sample center module; the sample center module is used to store and upload sample information of a batch of user test items, which can be added through template import or page editing, and specifically covers sample information such as the inspection batch, sample number, name, gender, pathological diagnosis and sampling site; in addition, the added sample information can also be edited and deleted.

[0112] In actual applications, the report generation subsystem also includes an experimental center module; the experimental center module is used to store and upload experimental information of a batch of users, which can be entered through template import, specifically covering experimental information such as inspection batch, sample number, DNA concentration, quality grade, pre-library concentration, total amount of pre-library, etc.; in addition, the added experimental information can also be entered and edited.

[0113] In actual applications, the report generation subsystem also includes an audit center module; the audit center module includes a first review and a second review, which calls the target database based on the individualized data provided by the user to review and interpret the detected genetic variations; the first review is used to review the quality control information of the individualized data, and to conduct a preliminary interpretation of the individualized somatic variations detected to match medication, prognosis or diagnosis information; the second review is used to test the results of the first review, ensure the accuracy of the data, and adjust and optimize the report format.

[0114] In actual applications, the report generation subsystem also includes a report center module; the report center module is used to download and manage tumor precision medication interpretation reports with individualized data; different report template logos, included data types, and layout formats are set for different users.

[0115] Individualized data is based on sample information and experimental information provided by the user, and tumor somatic variants detected after analysis by the reporting system. The review and interpretation is based on the quality control information of the detected tumor somatic variants on the reporting system interface, and at the same time, it interprets whether the detected tumor somatic variants correctly match the medication, prognosis and diagnosis information, and adjusts and optimizes the report in terms of content and format. The automated generation of tumor precision medication reports is to create a tumor NGS data analysis management system through programming, which links sample information, experimental information, detected variants, quality control information, medication instructions, prognosis and diagnosis information to automatically form a tumor precision medication report.

[0116] In summary, if the site information of the gene mutation is known, the present invention can search for the biological function of the mutation site through the gene mutation analysis module; further match the clinical information such as medication, prognosis and diagnosis of the mutation site through the key field label. That is, starting from the three dimensions of medication guidance, prognosis evaluation and auxiliary diagnosis, the tumor somatic cell mutation is evaluated in an all-round, multi-level, accurate and up-to-date clinical significance to the greatest extent; and further combined with the literature retrieval subsystem, database management subsystem and report generation subsystem to form a set of tumor somatic cell mutation interpretation database with interactive matching, which is suitable for the precision treatment of tumor patients.

[0117] The applicant declares that the above is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Those skilled in the art should understand that any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention shall fall within the protection scope and disclosure scope of the present invention.

Claims

1. A database for interpreting tumor somatic cell mutations, characterized in that: The interpretation database includes a document retrieval subsystem, a database management subsystem and a report generation subsystem; The literature retrieval subsystem is used to search and parse external databases online, and obtain the search and parsing results as evidence for gene variation, medication guidance, prognosis assessment and auxiliary diagnosis; The database management subsystem is used to integrate evidence of gene variation, medication guidance, prognosis assessment and auxiliary diagnosis, and to perform hierarchical classification and update management. The target database is divided to store tumor gene variation, targeted therapy, chemotherapy, immunotherapy, tumor prognosis and tumor diagnosis information respectively, and interactively associates by setting key field tags, and submits and enters into the reporting system; The report generation subsystem is used to call the target database according to the individualized data provided by the user, and automatically generate a tumor precision medication report after review and interpretation.

2. The interpretation database according to claim 1, characterized in that: The document retrieval subsystem includes: Literature retrieval module: used to search and parse external databases online, and obtain the search and parsing results as evidence for gene variation, medication guidance, prognosis assessment and auxiliary diagnosis; Gene mutation analysis module: used to analyze the target documents in the document retrieval module and obtain the analysis results about gene mutation; Targeted therapy analysis module: used to analyze the target literature in the literature retrieval module and obtain analysis results on targeted therapy; Chemical drug treatment analysis module: used to analyze the target documents in the literature search module and obtain analysis results on chemical drug treatment; Immunotherapy analysis module: used to analyze the target documents in the literature search module and obtain analysis results on immunotherapy; Prognosis analysis module: used to analyze the target literature in the literature retrieval module to obtain analysis results about tumor prognosis; Diagnostic analysis module: used to analyze the target documents in the document retrieval module and obtain analysis results about tumor diagnosis.

3. The interpretation database according to claim 1 or 2, characterized in that: The database management subsystem includes: Database architecture module: including gene variation analysis, targeted therapy and evidence level, chemotherapy and evidence level, immunotherapy and evidence level, prognosis and evidence level, and diagnosis and evidence level information; Key field labeling module: used to assign specific labels to variant sites according to the biological functions of gene variants, and to interactively associate with medication, prognosis and diagnostic target databases to form a tumor somatic variant interpretation database with interactive matching functions; Database classification module: used to classify the reviewed data according to the target database in the database architecture module and the somatic variation interpretation guideline standard, and store the classified target database to obtain a classified database; Database classification module: used to divide the target database in the database classification module into clinically significant groups, and store the classified target database to obtain a classified database; Database review module: used to send the data in the database architecture module sorted out after searching or updating by the searcher to the data administrator, so that the data administrator can review the target database; Database submission module: used to submit each target database in the database architecture module integrated by the data administrator to the report generation subsystem; Database update module: used to modify the information of each target database in the database architecture module and keep it dynamically updated to ensure the accuracy and timeliness of the interpretation database.

4. The interpretation database according to claim 3, characterized in that: The gene variation analysis is used to store information on the relationship between the biological function of genes and the occurrence and development of tumors; Preferably, the targeted therapy and evidence level are used to store targeted drugs and drug evidence level information related to tumor precision medication; Preferably, the chemotherapy treatment and evidence level are used to store chemotherapy drugs and drug evidence level information related to tumor precision medication; Preferably, the immunotherapy and evidence level are used to store information on immunotherapy drugs and drug evidence levels related to tumor precision medication; Preferably, the prognosis and evidence level are used to store tumor prognosis and prognosis evidence level information, and the diagnosis and evidence level are used to store tumor diagnosis and diagnosis evidence level information.

5. The interpretation database according to any one of claims 1 to 4, characterized in that: The report generation subsystem includes: Sample center module: used to store and upload sample information of user-submitted items; Experiment center module: used to store and upload user's experimental information; Review center module: used to call the target database based on the individualized data provided by the user, and to review and interpret gene mutations; Report center module: used to download and manage tumor precision medicine report forms with individualized data.

6. The interpretation database according to any one of claims 1 to 5, characterized in that: The information on tumor gene mutations in the target database includes genes, transcripts, nucleotide changes, amino acid changes, mutation types, mutation functions, mutation analysis and reference data; Preferably, the information on targeted therapy, chemotherapy and immunotherapy in the target database includes Chinese and English drug evidence, clinical trial numbers, clinical trial nicknames, references, cancer names in the evidence, Chinese and English drug names, major cancer types, evidence classification and drug efficacy data; Preferably, the tumor prognosis information in the target database includes genes, cancer names in evidence, prognosis analysis and reference data; Preferably, the tumor diagnosis information in the target database includes genes, cancer names in evidence, diagnostic analysis and reference data.

7. The interpretation database according to any one of claims 1 to 6, characterized in that: The individualized data is based on the sample information and experimental information uploaded by the user, and the tumor somatic mutations detected after analysis by the reporting system; Preferably, the review and interpretation is to review the quality control information of the detected tumor somatic variants based on the reporting system interface, and to interpret whether the detected tumor somatic variants correctly match the medication, prognosis and diagnosis information, and to adjust and optimize the report in terms of content and format; Preferably, the automated generation of a tumor precision medication report is accomplished by programming a tumor NGS data analysis and management system, which correlates sample information, experimental information, detected variations, quality control information, medication instructions, prognosis, and diagnosis information to automatically generate a tumor precision medication report.

8. A method for constructing an interpretation database according to any one of claims 1 to 7, characterized in that: The construction method comprises: (1) Online search and analysis of external databases to obtain search and analysis results as evidence for gene variation, medication guidance, prognosis assessment, and auxiliary diagnosis; (2) Integrate evidence on gene variation, medication guidance, prognosis assessment, and auxiliary diagnosis, and perform hierarchical classification and updated management; (3) Divide the target databases to store information on tumor gene mutations, targeted therapy, chemotherapy, immunotherapy, tumor prognosis, and tumor diagnosis, and interactively associate them by setting key field labels, submitting and entering them into the reporting system; (4) The reporting system calls the target database based on the individualized data provided by the user and automatically generates a tumor precision medication report after review and interpretation.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program or instructions, and when the computer program or instructions are executed by the processor, the steps in the construction method described in claim 8 are implemented.

10. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program or instructions, and the computer program or instructions, when executed by the processor, implement the steps in the construction method described in claim 8.

Images