Gene recognition method based on empirical Bayesian and Mendel randomization fusion

Abstract

The invention relates to a gene recognition method based on empirical Bayesian and Mendel randomization fusion. The method comprises the following steps: analyzing whole genome association analysis data by adopting empirical Bayesian meta-information to obtain an analysis result; correcting a statistical value of each SNP in the whole genome based on comprehensive hierarchical meta-information analysis of empirical Bayesian; integrating the whole genome association analysis data based on the Mendel randomization with the eQTL data and the mQTL data respectively, and obtaining a gene recognition result according to the overlapping part of the integration results of the whole genome association analysis data based on the Mendel randomization integrating with the eQTL data and the mQTL data.According to the method, the recognition speed of AD related genes can be greatly increased, existing data are fully utilized and the recognition speed of disease related genes is increased, and the research and development cost is saved; and the calculation result can screen out a great part of genes, so that a valuable research range is provided for subsequent biological experiments.

Description

Technical field [0001] The invention relates to the technical field of gene recognition, and is a gene recognition method based on the fusion of empirical Bayes and Mendelian randomization. Background technique [0002] At present, most scholars use methods such as developing related reagents and constructing biological experiments to identify genes related to Alzheimer's disease (AD). These methods are extremely time-consuming and extremely expensive. In recent years, with the continuous development of computer technology and the upgrading of sequencing technology, a large number of computer algorithms have been used to identify genes related to diseases. At present, using genome-wide association analysis (GWAS) to find risk genes for AD is a common method. At the beginning of this century, many research groups were identifying susceptibility sites for AD, but the actual results were not ideal. The only susceptibility gene found in different research groups was sorl1. The mai...

AI Technical Summary

Problems solved by technology

However, GWAS still has certain limitations

For example, the strategy is based on the hypothesis of "common diseases - common variants", which misses rare variants that may play a more important role in the etiology (minor allele frequency (MAF<0.005)); The associated SNP is not necessarily the real disease-causing locus, but only the "label" SNP that is LD-associated with the real disease-causing locus, especially some signals located in the so-called "desert area" of the gene, which are of great significance to the biological process of elucidating genetic variation. GWAS usually analyzes the marginal effect of a single locus according to the principle of the most obvious statistical difference, while ignoring the multi-gene interaction in complex diseases

Therefore, GWAS still cannot fully reveal the genetic susceptibility factors of complex diseases. It is only an important part of exploring the genetic etiology mechanism of complex diseases. How to dig deep into GWAS susceptibility loci and find out the real pathogenic loci, And exploring how these non-coding sequences exert biological mechanisms has become another challenge in genetics research

Images