Research On Epistasis Detection Method In Complex Diseases

With the rapid development of high-throughput technologies,it is possible to conduct association study of complex diseases in the genome-wide.Genome-wide association studies(GWAS)are the case-control studies that examine genetic variants,usually Single Nucleotide Polymorphisms(SNPs),in different individuals to see if any genetic variants are associated with complex diseases.Epistasis is defined as the interactions between two or more SNPs(or genes),it is one of the important genetic factor that affect complex diseases.Researchers have proposed a variety of methods to detect epistasis in genome-wide,but these methods still have some problems,such as low power and difficult to handle large data sets.This paper mainly studies the epistasis detection methods,including the following two points:(1)For existing entropy-based epistasis detection method doesn't consider data sets of disequilibrium and has low efficiency,this paper proposes a novel entropy-based method for detecting epistasis.The method betters take into account of the application in imbalanced datasets and add a penalty factor for the conditional entropy,which can better amplify the allele difference of SNP combination between case and control,so it is more likely to detect the SNP combination which is true associated with disease.A series experiments on the simulated datasets show that this method improves the detection efficiency.(2)When epistasis detection in genome-wide,the computational load is huge,so this paper proposes a novel two-stage method(TwoFC)conduct epistasis detection.The method fuse the G2 test and absolute probability difference function as a new scoring function to measure the strength of association between SNPs and disease status,the fused scoring function is an excellent measure of the strength of such association,so it is more likely to detect the SNP combination which is true associated with disease.In the first stage,TwoFC obtain a predefined number of candidate SNPs determined by the fused scores;in the second stage,TwoFC detects epistasis among the candidate SNPs.The results of experiments on the simulated data and a real disease data set show that TwoFC has better detection efficiency.