Tumor prognosis risk assessment system based on fusion bioinformatics method
By developing a tumor prognosis risk assessment system based on fusion bioinformatics methods, the problem of complex programming and version mismatch in tumor microenvironment research is solved, and efficient and accurate tumor prognosis risk assessment is achieved to assist clinical decision-making.
Patent Information
- Application Number
- CN202311742099.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-12-18
- Publication Date
- 2025-03-14
AI Technical Summary
The existing technology has complex programming process and version mismatch problems in the research of tumor microenvironment, which makes the code unable to run and it is difficult to quickly find tumor markers, affecting clinical diagnosis and treatment.
Develop a tumor prognostic risk assessment system based on fusion bioinformatics methods, including data acquisition module, biomarker acquisition module, model construction module, clinical risk assessment module and comprehensive risk assessment module. Through multivariate Cox regression analysis and mutual information correlation analysis methods of information entropy, the programming complexity is reduced and the accuracy of evaluation results is improved.
It greatly reduces time and learning costs, improves the accuracy and efficiency of tumor prognosis risk assessment, and assists doctors in making clinical decisions faster and better.
Smart Images
Figure CN119626525A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of tumor prognosis risk assessment, and in particular to a tumor prognosis risk assessment system based on a fusion bioinformatics method. Background Art
[0002] Tumor is one of the major diseases that threaten human health. Although targeted therapy and immunotherapy are superior to traditional surgery, radiotherapy and chemotherapy in specifically killing tumor cells, both treatment methods have certain limitations in their scope of application. For example, some patients cannot use corresponding targeted treatments due to the lack of effective therapeutic targets, or immune checkpoint inhibitors are ineffective due to the lack of sufficient immune infiltrating cells in tumor tissues. These limitations have led to huge challenges in the clinical treatment of tumors, and there are many problems that need to be solved urgently. Therefore, further developing accurate and effective tumor prevention, diagnosis and treatment methods is an important part of tumor research.
[0003] In recent years, with the great success of tumor immunotherapy, more and more studies have begun to focus on the tumor immune microenvironment, which provides important clues for deciphering the interaction between tumors and immune cells.
[0004] Immune cells infiltrating in tumor tissues play an important regulatory role in the development and evolution of tumors. For example, immune-infiltrating T cells account for a high proportion in the microenvironment of many types of tumors, such as skin melanoma, lung cancer, and colon cancer. They are the key to tumor immunotherapy, and their content has been shown to be closely related to the prognosis of tumor patients. Therefore, studying the tumor immune microenvironment not only helps to decipher the mechanism of tumor formation and progression, but also provides important clues and guidance for the development of new tumor immunotherapy methods.
[0005] In the field of bioinformatics, a relatively mature algorithm has been developed for the study of tumor microenvironment. For each stage of tumor microenvironment analysis, it is necessary to set up a working directory and install the corresponding package in the R environment. The process is cumbersome and complicated, and the code may not run due to version mismatch and other issues. In reality, it is clinicians or people engaged in basic medical research who can get close to tumor patients and collect relevant data. For those who have no programming experience but have the power to make decisions on clinical diagnosis and treatment, finding tumor markers can provide important clues and guidance for the diagnosis, treatment and rehabilitation of the disease. Summary of the invention
[0006] The present invention provides a tumor prognosis risk assessment system based on a fusion bioinformatics method to overcome the above technical problems.
[0007] A tumor prognosis risk assessment system based on a fusion bioinformatics method includes a data acquisition module, a biomarker acquisition module, a model building module, a clinical risk assessment module and a comprehensive risk assessment module.
[0008] The data acquisition module is used to obtain the whole transcriptome sequencing data set and clinical data set of the tumor.
[0009] The biomarker acquisition module is used to obtain biomarkers from the whole transcriptome sequencing data set based on bioinformatics methods, where biomarkers are genes related to diseases.
[0010] The model building module is used to build a high- and low-risk group model for evaluating prognosis through multivariate Cox regression analysis. The high- and low-risk group model for evaluating prognosis is used to perform multivariate Cox regression analysis on biomarkers and prognostic risks, and the relationship between biomarkers and prognostic risks is expressed according to the results of the multivariate Cox regression analysis.
[0011] The clinical risk assessment module is used to analyze the relationship between the high- and low-risk group models for prognosis assessment and the clinical data set based on the mutual information correlation analysis method of information entropy.
[0012] The comprehensive risk assessment module is used to determine the prognostic risk of a tumor based on the relationship between biomarkers and prognostic risk, and the relationship between prognostic risk and clinical data sets.
[0013] Preferably, the bioinformatics-based method for obtaining biomarkers from a whole transcriptome sequencing data set includes finding differentially expressed genes from a whole transcriptome sequencing data set by a differential analysis method; then using the differentially expressed genes to perform GO and KEGG functional pathway enrichment analysis, wherein the GO and KEGG functional pathway enrichment analysis is used to verify the biological processes and signal pathways in which the differentially expressed genes participate; finally, constructing a gene co-expression network, calculating the number of gene connection edges in the co-expression network, determining the importance of the gene based on the number of gene connection edges, and determining whether the gene is a biomarker based on the importance of the gene.
[0014] Preferably, the mutual information correlation analysis method based on information entropy analyzes the relationship between the high-risk and low-risk group models for evaluating prognosis and the clinical data set, including:
[0015] Step 1: Calculate the information entropy of high and low disease risk groups;
[0016] H(high risk) = -p(high risk)logp(high risk)
[0017] H(low risk) = -p(low risk)logp(low risk)
[0018] Step 2: Calculate the conditional entropy of high and low disease risk groups under different clinical traits in the clinical data set;
[0019]
[0020]
[0021] Step 3: Calculate the mutual information value between high-risk and low-risk groups and clinical traits;
[0022] I(high risk, clinical trait) = H(high risk) - H(high risk | a certain clinical trait) I(low risk, clinical trait) = H(low risk) - H(low risk | a certain clinical trait).
[0023] Preferably, the step of finding differentially expressed genes from the whole transcriptome sequencing data set is to obtain high and low gene expression groups from the whole transcriptome sequencing data set, and perform differential expression analysis on the patient grouping expression spectrum matrix obtained by the high and low gene expression groupings using the "limma" package in the R software.
[0024] Preferably, the GO and KEGG functional pathway enrichment analysis is to use the R packages "colorspace", "stringi", and "ggplot2" to perform GO and KEGG functional pathway enrichment analysis on the differentially expressed genes.
[0025] The present invention provides a tumor prognosis risk assessment system based on the fusion bioinformatics method, which redevelops the currently more mature tumor marker bioinformatics analysis method, transforms the complex programming process into a tumor prognosis risk assessment system, and greatly reduces the time cost and learning cost. In addition, in order to assist doctors in making clinical decisions faster and better, an improved mutual information correlation analysis method based on information entropy is used to analyze the correlation between high-risk and low-risk groups and clinical traits, reducing the deviation caused by the small sample size and improving the accuracy of the assessment results. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0027] Figure 1 It is a system structure diagram of the present invention;
[0028] Figure 2 It is a visualized bioinformatics analysis platform of the present invention. DETAILED DESCRIPTION
[0029] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0030] Figure 1 It is the system structure diagram of the present invention, such as Figure 1 As shown, the method of this embodiment may include:
[0031] A tumor prognosis risk assessment system based on a fusion bioinformatics method includes a data acquisition module, a biomarker acquisition module, a model building module, a clinical risk assessment module and a comprehensive risk assessment module.
[0032] The data acquisition module is used to obtain the whole transcriptome sequencing data set and clinical data set of the tumor.
[0033] The biomarker acquisition module is used to obtain biomarkers from the whole transcriptome sequencing data set based on bioinformatics methods, where biomarkers are genes related to diseases.
[0034] The model building module is used to build a high- and low-risk group model for evaluating prognosis through multivariate Cox regression analysis. The high- and low-risk group model for evaluating prognosis is used to perform multivariate Cox regression analysis on biomarkers and prognostic risks, and the relationship between biomarkers and prognostic risks is expressed according to the results of the multivariate Cox regression analysis.
[0035] The clinical risk assessment module is used to analyze the relationship between the high- and low-risk group models for prognosis assessment and the clinical data set based on the mutual information correlation analysis method of information entropy.
[0036] The comprehensive risk assessment module is used to determine the prognostic risk of a tumor based on the relationship between biomarkers and prognostic risk, and the relationship between prognostic risk and clinical data sets.
[0037] Based on the above scheme, the present invention provides a tumor prognosis risk assessment system based on the fusion bioinformatics method, which redevelops the currently more mature tumor marker bioinformatics analysis method, transforms the complex programming process into a tumor prognosis risk assessment system, and greatly reduces the time cost and learning cost. In addition, in order to assist doctors in making clinical decisions faster and better, an improved mutual information correlation analysis method based on information entropy is used to analyze the correlation between high-risk and low-risk groups and clinical traits, reducing the deviation caused by the small sample size and improving the accuracy of the assessment results.
[0038] Specifically, this embodiment provides the details of the prognosis evaluation system, including
[0039] A tumor prognosis risk assessment system based on a fusion bioinformatics method includes a data acquisition module, a biomarker acquisition module, a model building module, a clinical risk assessment module and a comprehensive risk assessment module.
[0040] The data acquisition module is used to obtain the whole transcriptome sequencing data set and clinical data set of the tumor. The transcription spectrum expression data and clinical pathological information of one or more diseases are selected in the TCGA database, and the transcription spectrum expression data is used as the whole transcriptome sequencing data set. The whole transcriptome sequencing data set includes the TCGA-STAD data set, which contains multiple tumor tissue samples and multiple adjacent tissue samples. The data set includes two files, one is a meta file and the other is a cart file. Among them, the meta file stores the correspondence between the sample file name (file_name) and the sample (entity_submitter_id), and the cart file stores the expression data of each sample. By running the Perl language script named merge.pl, the sample data preprocessing can be completed to obtain the matrix data of "row (gene name) * column (sample number)".
[0041] Clinical pathological information is used as a clinical data set. The traits in the clinical data set include age, gender, tumor stage, malignancy, etc.
[0042] The biomarker acquisition module is used to acquire biomarkers from a whole transcriptome sequencing data set based on a bioinformatics method, wherein the biomarker is a gene associated with a disease, and the acquisition of biomarkers from a whole transcriptome sequencing data set based on a bioinformatics method includes finding differentially expressed genes from the whole transcriptome sequencing data set by a differential analysis method; finding differentially expressed genes from the whole transcriptome sequencing data set is to obtain high and low gene expression groupings from the whole transcriptome sequencing data set, and perform differential expression analysis on the patient grouping expression spectrum matrix obtained by the high and low gene expression groupings through the "limma" package in the R software.
[0043] Then, the differentially expressed genes were used to perform GO and KEGG functional pathway enrichment analysis, which was used to verify the biological processes and signal pathways in which the differentially expressed genes were involved; the GO and KEGG functional pathway enrichment analysis was performed using the R packages "colorspace", "stringi", and "ggplot2" to perform GO and KEGG functional pathway enrichment analysis on the differentially expressed genes. Finally, a gene co-expression network was constructed, the number of gene connection edges in the co-expression network was calculated, the importance of the gene was determined based on the number of gene connection edges, and whether the gene was a biomarker was determined based on the importance of the gene.
[0044] Based on the patient's transcriptional profile information, the R package "estimate" can also be used to evaluate the tumor microenvironment-related indicators Stromal Score, Immune Score, and ESTIMATE Score.
[0045] Specifically, the currently disclosed methods for biomarker identification for cancer treatment and diagnosis are used in combination to fully analyze transcriptome data and clinical data, dig out disease biomarkers, and thus establish a high- and low-risk group model for assessing prognosis, and find out the correlation between prognostic effect and clinical characteristics. Researchers only need to select the transcriptional spectrum expression data and clinical pathological information of any disease in the TCGA database, and use integrated bioinformatics-related methods to quickly find the biomarkers of the disease, and verify the results through literature retrieval to ensure the effectiveness of the method.
[0046] The model building module is used to build a high- and low-risk group model for evaluating prognosis through multivariate Cox regression analysis. The high- and low-risk group model for evaluating prognosis is used to perform multivariate Cox regression analysis on biomarkers and prognostic risks, and the relationship between biomarkers and prognostic risks is expressed according to the results of the multivariate Cox regression analysis.
[0047] The clinical risk assessment module is used to analyze the relationship between the high- and low-risk group models for prognosis assessment and the clinical data set based on the mutual information correlation analysis method of information entropy.
[0048] The mutual information correlation analysis method based on information entropy analyzes the relationship between the high-risk and low-risk group models for evaluating prognosis and the clinical data set, including:
[0049] Step 1: Calculate the information entropy of high and low disease risk groups;
[0050] H(high risk) = -p(high risk)logp(high risk)
[0051] H(low risk) = -p(low risk)logp(low risk)
[0052] Step 2: Calculate the conditional entropy of high and low disease risk groups under different clinical traits in the clinical data set;
[0053]
[0054]
[0055] Step 3: Calculate the mutual information value between high-risk and low-risk groups and clinical traits;
[0056] I(high risk, clinical trait) = H(high risk) - H(high risk | a certain clinical trait) I(low risk, clinical trait) = H(low risk) - H(low risk | a certain clinical trait).
[0057] The comprehensive risk assessment module is used to determine the prognostic risk of a tumor based on the relationship between biomarkers and prognostic risk, and the relationship between prognostic risk and clinical data sets.
[0058] Specifically, this embodiment constructs the tumor prognosis risk assessment model as a visual bioinformatics analysis platform, such as Figure 2 As shown, the following steps are included:
[0059] Step 1: Cross-language script encapsulation: Encapsulate the Perl and R language codes in the bioinformatics analysis process and make them executable in the command prompt CMD in the form of a script.
[0060] Step 2: Package the R packages to be used in tumor microenvironment analysis, differential gene analysis, and enrichment analysis to facilitate unified versions and avoid problems such as program failure or errors due to version mismatch.
[0061] Step 3: Design the functional modules of the GUI interface. Each important step in biomarker identification is designed as a function button. The biomarker identification process is mainly divided into data set preprocessing, tumor microenvironment index analysis, screening of differentially expressed genes and gene enrichment analysis.
[0062] Step 4: Implement dataset preprocessing. Put the packaged data preprocessing script merge.pl and the dataset downloaded from the TCGA database in the same folder. During the Java GUI development process, call the Runtime.getRuntime() module, and obtain the directory where the user dataset is located by calling JFileChooser. Use command line statements to switch the working directory and run the Perl script to implement dataset preprocessing.
[0063] Step 5: Implementation of the tumor microenvironment index analysis module. The R package "estimate" was used to evaluate the tumor microenvironment related indicators Stromal Score, Immune Score, and ESTIMATE Score.
[0064] Step 6: Implementation of screening differentially expressed gene modules. The R package "limma" was used to perform differential expression analysis on the patient group expression spectrum matrix obtained by grouping high and low gene expression levels.
[0065] Step 7: Implementation of gene enrichment analysis module. GO and KEGG functional enrichment analysis of differentially expressed genes was performed using the R package "ggplot2". Among them, steps 5, 6, and 7 can be summarized as downloading disease-related whole transcriptome sequencing datasets from the TCGA database and using the ESTIMATE algorithm to evaluate the score of immune infiltrating cells. Differentially expressed genes (DEGs) analysis was performed using the R software package, a series of enrichment analyses were performed using the Metascape tool, and hierarchical clustering and survival analysis were performed on the enrichment analysis results.
[0066] Specifically, the prediction results of the evaluation system of this embodiment can be further verified by literature retrieval, including
[0067] Step 1: The differentially expressed genes obtained by bioinformatics methods are used to form a "gene name" keyword list. The literature is searched using crawler technology in the PubMed module of NCBI to obtain a list of literature related to each biomarker.
[0068] Step 2: Combined with the disease type and differentially expressed genes, the keywords are "disease name gene name", and continue to use crawler technology to search for literature in the PubMed module of NCBI to obtain a list of all research literature on each biomarker and related diseases.
[0069] Step 3: Analyze the literature results retrieved by using "disease name gene name" as the keyword. If there are a large number of literature, it means that there are many experiments or research analyses on the biomarker of the disease, so researchers can reduce the experimental research on the biomarker; if there are a small number of literature, it means that there are not many studies on the biomarker of the disease, and the literature of the biomarker retrieved in step 1 can be used to study it to understand its main function and the biological process involved. In addition, clinical data can be collected and further related clinical experiments can be conducted to analyze and explore its role in the occurrence and development of the disease.
[0070] Overall beneficial effects:
[0071] The present invention provides a tumor prognosis risk assessment system based on the fusion bioinformatics method, which redevelops the currently more mature tumor marker bioinformatics analysis method, transforms the complex programming process into a tumor prognosis risk assessment system, and greatly reduces the time cost and learning cost. In addition, in order to assist doctors in making clinical decisions faster and better, an improved mutual information correlation analysis method based on information entropy is used to analyze the correlation between high-risk and low-risk groups and clinical traits, reducing the deviation caused by the small sample size and improving the accuracy of the assessment results.
[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A tumor prognosis risk assessment system based on a fusion bioinformatics method, characterized in that: It includes data acquisition module, biomarker acquisition module, model building module, clinical risk assessment module and comprehensive risk assessment module. The data acquisition module is used to obtain the whole transcriptome sequencing data set and clinical data set of the tumor. The biomarker acquisition module is used to obtain biomarkers from the whole transcriptome sequencing data set based on bioinformatics methods, where biomarkers are genes related to diseases. The model building module is used to build a high- and low-risk group model for evaluating prognosis through multivariate Cox regression analysis. The high- and low-risk group model for evaluating prognosis is used to perform multivariate Cox regression analysis on biomarkers and prognostic risks, and the relationship between biomarkers and prognostic risks is expressed according to the results of the multivariate Cox regression analysis. The clinical risk assessment module is used to analyze the relationship between the high- and low-risk group models for prognosis assessment and the clinical data set based on the mutual information correlation analysis method of information entropy. The comprehensive risk assessment module is used to determine the prognostic risk of a tumor based on the relationship between biomarkers and prognostic risk, and the relationship between prognostic risk and clinical data sets.
2. A tumor prognosis risk assessment system based on a fusion bioinformatics method according to claim 1, characterized in that: The method of obtaining biomarkers from a whole transcriptome sequencing data set based on a bioinformatics method includes finding differentially expressed genes from a whole transcriptome sequencing data set by a differential analysis method; then using the differentially expressed genes to perform GO and KEGG functional pathway enrichment analysis, wherein the GO and KEGG functional pathway enrichment analysis is used to verify the biological processes and signal pathways in which the differentially expressed genes participate; finally, constructing a gene co-expression network, calculating the number of gene connection edges in the co-expression network, determining the importance of the gene according to the number of gene connection edges, and determining whether the gene is a biomarker according to the importance of the gene.
3. The tumor prognosis risk assessment system based on the fusion bioinformatics method according to claim 1, characterized in that: The mutual information correlation analysis method based on information entropy analyzes the relationship between the high-risk and low-risk group models for evaluating prognosis and the clinical data set, including: Step 1: Calculate the information entropy of high and low disease risk groups; H(high risk) = -p(high risk)logp(high risk) H(low risk) = -p(low risk)logp(low risk) Step 2: Calculate the conditional entropy of high and low disease risk groups under different clinical traits in the clinical data set; Step 3: Calculate the mutual information value between high-risk and low-risk groups and clinical traits; I(high risk, clinical trait) = H(high risk) - H(high risk|certain clinical trait) I(low risk, clinical trait)=H(low risk)-H(low risk|certain clinical trait).
4. The tumor prognosis risk assessment system based on the fusion bioinformatics method according to claim 2, characterized in that: The method of finding differentially expressed genes from the whole transcriptome sequencing data set is to obtain high and low gene expression groups from the whole transcriptome sequencing data set, and perform differential expression analysis on the patient grouping expression spectrum matrix obtained by the high and low gene expression grouping using the "limma" package in the R software.
5. The tumor prognosis risk assessment system based on the fusion bioinformatics method according to claim 2, characterized in that: The GO and KEGG functional pathway enrichment analysis is to use the R packages "colorspace", "stringi", and "ggplot2" to perform GO and KEGG functional pathway enrichment analysis on the differentially expressed genes.