Exome sequencing of 20,791 cases of type 2 diabetes and 24,440 controls

Abstract

By identifying molecular alterations causally associated with human traits, studies of naturally occurring genetic variation can yield crucial clues about disease pathogenesis. Protein-coding variants that strongly affect disease risk are of particular value, as they directly implicate specific genes. Here, through a large-scale exome sequence analysis, we investigate the role of coding variation in the genetic basis and biology of type 2 diabetes (T2D). Our results identify four gene-level associations at exome-wide significance and suggest that rare coding variant T2D associations are commonplace but contribute minimally to disease heritability. Several candidate T2D-relevant gene sets – including established T2D drug targets – demonstrate set-level evidences of association, but we estimate gene-specific signals within them will not achieve exome-wide significance until at least ten-fold more samples are available. We propose a method to interpret these modest rare-variant associations and incorporate them into target or gene prioritization efforts. Our data are freely available for analysis at www.type2diabetesgenetics.org.