Gene mutation pathogenicity detection method and system based on neural network and medium

Abstract

The invention discloses a gene mutation pathogenicity detection method and system based on a neural network and a medium. The method comprises the steps of inputting a gene detection VCF file to be detected and an HPO phenotype; obtaining a characteristic value of each gene variation according to the to-be-detected gene detection VCF file and the HPO phenotype; for each gene variation, inputting the characteristic value of the gene variation into a trained neural network model to obtain a pathogenicity comprehensive analysis result of the gene variation, wherein the neural network model is pre-trained to establish a mapping relationship between the characteristic value of each gene variation and the pathogenicity comprehensive analysis result of each gene variation. According to the invention, the subjective defect of manual analysis is overcome, and various factors influencing the pathogenicity of gene mutation can be comprehensively considered, so that the comprehensive analysis result is more objective and effective, the accuracy of pathogenicity analysis of gene mutation is greatly improved, and the gene interpretation efficiency is improved.

Description

technical field [0001] The gene detection and gene interpretation technology involved in the present invention specifically relates to a neural network-based gene mutation pathogenicity detection method, system and medium. Background technique [0002] In genetic testing, how to obtain the disease-causing gene mutation from thousands of gene mutations is the key to gene interpretation. The main methods of traditional gene mutation pathogenicity analysis include: (1) By searching professional databases in the biomedical field to see whether the corresponding gene or mutation has a pathogenic database record, these databases include OMIM, Orphanet, HGMD, Clinvar, etc. (2) Determine whether the gene mutation is at risk through protein function prediction, commonly used protein function prediction software such as SIFT, PolyPhen2, MutationTaster, DANN, CADD, etc.; (3) According to the frequency of the gene mutation in the population, determine The risk of gene mutation, commonl...

AI Technical Summary

Problems solved by technology

The biggest defect of this method is that less pathogenic factors are considered, and some factors (such as mouse and zebrafish data) have little effect on human genetic diseases, and some key factors are not taken into account, resulting in acquired pathogenicity. The accuracy of comprehensive analysis value is not high

Images