A1017
Title: Bayesian fine mapping of phenome-wide transcriptome-wide association studies
Authors: Tianzhou Ma - University of Maryland (United States) [presenting]
Abstract: Transcriptome-wide association studies (TWAS) integrate genome-wide association studies and reference expression quantitative trait loci studies to identify genes with genetically predicted expression associated with a complex trait. Fine mapping methods have been developed to identify the potentially causal genes in the TWAS associated regions, possibly pointing to the molecular mechanism behind the association. The expansion of phenomic data like those from electronic health records (EHRs) in the past decade(s) has shifted the field from single trait studies to phenome-wide studies where a wide range of closely related phenotypes are jointly analyzed and how they are genetically regulated is studied. The focus is on a new Bayesian fine mapping method, FM-GPT (fine-mapping of causal genes for phenome-wide transcriptome-wide association studies), to identify causal genes for a multitude of phenotypes with possibly mixed data types in phenome-wide transcriptome-wide association studies. It is shown in simulations that FM-GPT was more powerful than other single-trait and multi-trait fine mapping methods in identifying the true set of causal genes without losing the false discovery rate control. The method is applied to two real data examples with multiple complex traits from UK Biobank and identified critical genes that potentially regulated subregional cortical thickness and clinical phenotypes spanning multiple systems.
