Search

CN-122024853-A - Method for analyzing prognostic gene characteristics of glioma based on machine learning

CN122024853ACN 122024853 ACN122024853 ACN 122024853ACN-122024853-A

Abstract

The invention relates to the field of biological detection and discloses an analysis method for prognostic gene characteristics of glioma based on machine learning, which comprises the steps of obtaining prognostic gene expression data and pathological variable data corresponding to glioma evaluation samples; and analyzing the prognosis risk of the glioma assessment sample according to the prognosis gene score and the pathological variable data to obtain a prognosis risk analysis result. The invention solves the problems of insufficient accuracy and lack of quantitative analysis model of the existing glioma prognosis evaluation method.

Inventors

  • Ma Jiangchun

Assignees

  • 浙江医院

Dates

Publication Date
20260512
Application Date
20260202

Claims (10)

  1. 1. A method for machine learning-based analysis of prognostic gene signatures of gliomas, the method comprising: Obtaining prognostic gene expression data and pathological variable data corresponding to glioma evaluation samples; determining a prognostic gene score for the prognostic gene expression data based on a pre-constructed prognostic scoring model; And analyzing the prognosis risk of the glioma assessment sample according to the prognosis gene scores and the pathological variable data to obtain a prognosis risk analysis result.
  2. 2. The method of claim 1, wherein analyzing the prognostic risk of the glioma assessment sample based on the prognostic gene score and the pathological variable data to obtain a prognostic risk analysis result comprises: Determining a risk grade group to which the glioma assessment sample belongs according to the prognostic gene score; Substituting the prognostic gene scores and the pathological variable data into a pre-constructed nomogram to obtain a prognostic state value of the glioma evaluation sample at a designated time point; And generating a prognosis risk analysis result of the glioma assessment sample based on the risk level group and the prognosis state value.
  3. 3. The method according to claim 2, wherein the constructing method of the pre-constructed nomogram includes: obtaining prognosis gene scores and corresponding pathological variable data of glioma training samples; carrying out regression analysis on the prognosis gene score and the pathological variable data to generate a correlation model; and constructing an alignment chart mapping the prognostic gene scores and the pathological variable data into the prognostic state values according to the preset weights of the variables in the association model.
  4. 4. The method of claim 1, wherein after determining the risk level group to which the glioma assessment sample belongs according to the prognostic gene score, the method further comprises: acquiring microenvironment characteristic data and/or tumor mutation load data of each risk grade group; taking the micro-environment characteristic data and/or the mutation load data as biological characteristic data, and analyzing a biological characteristic association relationship between the biological characteristic data and the risk grade group; The prognostic risk analysis result is generated based on the risk level groupings, the prognostic status values, and the biological feature associations.
  5. 5. The method of claim 1, wherein the method of constructing the prognostic scoring model comprises: obtaining gene expression data of glioma training samples; extracting expression data of copper death related genes from the gene expression data, and performing cluster analysis based on the expression data of the copper death related genes to obtain a plurality of molecular subtypes; Identifying differentially expressed genes between different of the molecular subtypes, and constructing the prognostic scoring model based on the differentially expressed genes.
  6. 6. The method of claim 5, wherein performing cluster analysis based on the expression data of the copper death-related gene to obtain a plurality of molecular subtypes comprises: performing consensus cluster analysis on the expression data of the copper death related genes to obtain target cluster numbers; Dividing the glioma training samples into a plurality of molecular subtypes according to the target cluster number.
  7. 7. The method of claim 5, wherein said constructing said prognostic scoring model based on said differentially expressed genes comprises: screening prognosis candidate genes related to the evaluation time points of the glioma training samples from the differential expression genes; performing feature selection on the prognosis candidate genes to obtain a prognosis feature gene set; The prognostic scoring model is constructed based on the expression levels of the individual genes in the prognostic signature set of genes.
  8. 8. An apparatus for analyzing prognostic gene signatures of gliomas based on machine learning, the apparatus comprising: The acquisition module is used for acquiring prognostic gene expression data and pathological variable data corresponding to the glioma evaluation sample; a calculation module for calculating a prognostic gene score of the prognostic gene expression data based on a pre-constructed prognostic score model; and the analysis module is used for analyzing the prognosis risk of the glioma assessment sample according to the prognosis gene scores and the pathological variable data to obtain a prognosis risk analysis result.
  9. 9. A computer device, comprising: a memory and a processor in communication with each other, the memory having stored therein computer instructions which, upon execution, cause the processor to perform the method of any of claims 1 to 7.
  10. 10. A computer readable storage medium having stored thereon computer instructions for causing a computer to perform the method of any one of claims 1 to 7.

Description

Method for analyzing prognostic gene characteristics of glioma based on machine learning Technical Field The invention relates to the field of biological detection, in particular to an analysis method for prognosis gene characteristics of glioma based on machine learning. Background Gliomas are the most common primary malignancy of the central nervous system, have a high degree of heterogeneity and invasiveness, and are generally worse prognosis, especially high-grade gliomas. Currently, prognosis evaluation of gliomas in clinical practice is mainly dependent on histopathological classification and limited molecular markers (e.g. IDH mutation, 1p/19q co-deletion status, etc.). However, these traditional indexes still have difficulty in realizing accurate personalized prognosis layering, and cannot fully reflect biological heterogeneity inside tumor and complex interaction with tumor microenvironment, so that accuracy of prognosis prediction needs to be improved. The deficiency and disadvantage of the prior art is that a method capable of integrating specific biological process (such as copper death) related gene expression information and clinical pathological variables and automatically and quantitatively evaluating glioma sample prognosis based on a data analysis model is lacking. This results in insufficient accuracy and objectivity of the prognosis evaluation, and it is difficult to meet the clinical demands for personalized, accurate prognosis judgment. Disclosure of Invention Therefore, the embodiment of the invention provides an analysis method for the prognosis gene characteristics of glioma based on machine learning, so as to solve the problems of insufficient accuracy and lack of quantitative analysis model of the existing glioma prognosis evaluation method. In a first aspect, embodiments of the present invention provide a method for analyzing prognostic gene signatures of gliomas based on machine learning, the method comprising: Obtaining prognostic gene expression data and pathological variable data corresponding to glioma evaluation samples; determining a prognostic gene score for the prognostic gene expression data based on a pre-constructed prognostic scoring model; And analyzing the prognosis risk of the glioma assessment sample according to the prognosis gene scores and the pathological variable data to obtain a prognosis risk analysis result. Further, the analyzing the prognosis risk of the glioma assessment sample according to the prognosis gene score and the pathological variable data to obtain a prognosis risk analysis result includes: Determining a risk grade group to which the glioma assessment sample belongs according to the prognostic gene score; Substituting the prognostic gene scores and the pathological variable data into a pre-constructed nomogram to obtain a prognostic state value of the glioma evaluation sample at a designated time point; And generating a prognosis risk analysis result of the glioma assessment sample based on the risk level group and the prognosis state value. Further, the method for constructing the pre-constructed nomogram comprises the following steps: obtaining prognosis gene scores and corresponding pathological variable data of glioma training samples; carrying out regression analysis on the prognosis gene score and the pathological variable data to generate a correlation model; and constructing an alignment chart mapping the prognostic gene scores and the pathological variable data into the prognostic state values according to the preset weights of the variables in the association model. Further, after determining the risk level group to which the glioma assessment sample belongs according to the prognostic gene score, the method further comprises: acquiring microenvironment characteristic data and/or tumor mutation load data of each risk grade group; taking the micro-environment characteristic data and/or the mutation load data as biological characteristic data, and analyzing a biological characteristic association relationship between the biological characteristic data and the risk grade group; The prognostic risk analysis result is generated based on the risk level groupings, the prognostic status values, and the biological feature associations. Further, the construction method of the prognosis score model comprises the following steps: obtaining gene expression data of glioma training samples; extracting expression data of copper death related genes from the gene expression data, and performing cluster analysis based on the expression data of the copper death related genes to obtain a plurality of molecular subtypes; Identifying differentially expressed genes between different of the molecular subtypes, and constructing the prognostic scoring model based on the differentially expressed genes. Further, the clustering analysis is performed based on the expression data of the copper death-related gene to obtain a plurality of molecular subtypes, includ