Abstract
AbstractBackgroundCharacterizing genetic effect heterogeneity across subpopulations with different environmental exposures is useful for identifying exposure-specific pathways, understanding biological mechanisms underlying disease heterogeneity and further pinpointing modifiable risk factors for disease prevention and management. Classical gene-by-environment interaction (GxE) analysis can be used to characterize genetic effect heterogeneity. However, it can have a high multiple testing burden in the context of genome-wide association studies (GWAS) and requires a large sample size to achieve sufficient power.MethodsWe adapt a colocalization method, SharePro, to account for effect heterogeneity in finemapping and subsequently improve power for GxE analysis. Through joint fine-mapping of exposure stratified GWAS summary statistics, SharePro can greatly reduce multiple testing burden in GxE analysis.ResultsThrough extensive simulation studies, we demonstrated that accounting for effect heterogeneity can improve power for fine-mapping. With efficient joint fine-mapping of exposure stratified GWAS summary statistics, SharePro alleviated multiple testing burden in GxE analysis and demonstrated improved power with well-controlled false discovery rate. Through analyses of smoking status stratified GWAS summary statistics, we identified genetic effects on lung function modulated by smoking status mapped to the genesCHRNA3,ADAM19andUBR1. Additionally, using sex stratified GWAS summary statistics, we characterized sex differentiated genetic effects on fat distribution and provided biologically plausible candidates for functional follow-up studies.ConclusionsWe have developed an analytical framework to account for effect heterogeneity in finemapping and subsequently improve power for GxE analysis. The SharePro software for GxE analysis is openly available athttps://github.com/zhwm/SharePro_gxe.
Publisher
Cold Spring Harbor Laboratory