Font Size: a A A

Research On Machine Learning Algorithm For Gene Analysis

Posted on:2021-11-17Degree:MasterType:Thesis
Country:ChinaCandidate:Y M CaoFull Text:PDF
GTID:2480306602976769Subject:Mathematics
Abstract/Summary:PDF Full Text Request
The association analysis of multi-site combination of genes and phenotype is a research hotspot in bioinformatics.Recent studies have shown that in the study of complex genetic diseases,compared with single gene loci,the association of multiple loci of multiple genes with phenotype is more significant.Glaucoma is the second largest irreversible blindness in the world,and high intraocular pressure is the main indicator of glaucoma.Therefore,studying the correlation between multi-site combinations and intraocular pressure is of great significance for studying the pathogenesis of glaucoma.IOP data is a continuous phenotype,but the existing methods are only suitable for analyzing the association of multisite combinations with discrete phenotypes,but not the association with continuous phenotypes.This paper first designed an ant colony algorithm to study the association analysis between two-point combination of genes and continuous phenotype.On a dataset of genes and intraocular pressure released by the European Bioinformatics Institute,we evaluated the performance of the design method in this paper.The experimental results show that the ant colony algorithm designed in this paper can find the two-point combination of genes that are significantly related to intraocular pressure,which provides new clues for studying the pathogenesis of glaucoma.In addition,the ant colony algorithm designed in this paper provides a new idea for studying the association between multi-site combinations and continuous phenotypes of other diseases.This paper further studies the association analysis between the combination of three loci and continuous phenotype.First,a deep neural network is used to predict the effect of unit points of non-coding DNA segments on chromosomal functional units,and the most significant unit points are selected on the glaucoma intraocular pressure data set studied in this paper.Furthermore,these unit points are combined with the two-site combination selected by the ant colony algorithm to form a three-site combination,and the relationship between these three-site combinations and intraocular pressure is studied.The experimental results show that the three-point combination selected by the above method has a more significant correlation with the intraocular pressure data than the unit point and two-point combination.This research method provides a new research idea for studying the association between multi-site combinations of genetic data and continuous phenotypes.
Keywords/Search Tags:Machine learning, Ant colony algorithm(ACO), Artificial neural network, Genome-wide association study(GWAS), Intelligent diagnosis of glaucoma, Regression analysis of genes and traits, Non-coding region genes, Gene interaction
PDF Full Text Request
Related items