A Large-Scale Genome-Wide Enrichment Analysis Identifies New Trait-Associated Genes, Pathways and Tissues Across 31 Human Phenotypes

• Activity Number: 667 - Statistical Genetics
• Type: Contributed
• Date/Time: Thursday, August 3, 2017 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): Xiang Zhu* and Matthew Stephens
• Companies: University of Chicago and University of Chicago
• Keywords: Summary statistics ; Bayesian regression ; Genetics ; Association study ; Gene set enrichment analysis ; Variational Bayes

Abstract

Genome-wide association studies (GWAS) aim to identify genetic factors that are associated with complex traits. However, individual genetic variants have small effects, making them hard to identify. In addition, lists of individual variant associations give limited biological insights. "Enrichment analyses" can address these problems by focussing on biological pathways, instead of individual genetic variants. Here we develop an efficient enrichment analysis method that jointly models GWAS summary statistics at millions of variants, and use it to analyse 3,913 biological pathways and 64 tissue-based gene sets in 31 human phenotypes. Our results highlight several novel pathway and tissue associations. For example, the endochondral ossification pathway is enriched for associations with height, and liver-related genes are enriched for Alzheimer's disease. A key feature of our method is that inferred enrichment automatically informs new trait-associated genes. For example, enrichment in lipid transport genes suggests strong evidence for association between MTTP and low-density lipoprotein levels, whereas conventional analyses of the same data found no significant variant near this gene.