Pathway-based genome-wide association study using functional SNP effects

Abstract

Current single SNP based GWAS are insufficient to dissect the origin of the complex diseases. Recent studies apply pathway-based analysis as a new paradigm for GWAS to overcome their limitations, but there still exist unavoidable generation of the false positive noises according to the characteristic of LD block. In this study, we propose the way to eliminate the false noises and provide more reliable method to estimate the altered pathways, through considering functional deleterious SNP effects. Only consider functionally significant SNPs as the candidates of the pathway-based analysis, the resulted pathways reflect truer altered activity than the current approaches. We verify this result quantitatively through the classification, and show that our proposed approach has better classification performance than the conventional methods even though the current application coverage is limited.