Manuel A. Rivas, Guhan Venkataraman, Kai Yuan, John D. Rioux, Carl A. Anderson, Mark J. Daly, Ramnik J. Xavier, Dermot P.B. McGovern, Christine Stevens, Aleksejs Sazonovs, and Hailiang Huang
Background Genome-wide association studies (GWASs) have identified >200 genomic regions associated with IBD. These signals are mainly driven by common variants in noncoding regions, making it a challenge to extract biological insight. Targeted sequencing of genes in these regions has successfully identified putative causal variants, often rarer and independent of the initial GWAS hit. These coding variants have led to more direct functional experiments demonstrating causal mechanisms for at least ten genes. Methods To further rare variant-based discovery, we pursued large-scale exome sequencing of more than 30,000 Crohn's disease (CD) cases and population controls from more than 20 centers in the International IBD Genetics Consortium. Sequencing was performed at the Broad Institute using both Nextera (11,125 CD cases) and TWIST (7,442 CD cases) exome captures - and at the Sanger Institute (whole genomes 6,404 CD cases and Agilent exome 3,848 CD cases). Further replication was performed with a sample of 4,359 cases from Kiel sequenced at Regeneron. In each technical experiment, a comparable or larger number of sequenced controls were available for genetic association analysis. Results We conducted meta-analysis of two exome captures at the Broad Institute. To expand on GWAS, we focused on rare and low-frequency coding variants between 0.01% and 10% and estimated roughly 85% of all protein coding variants in this frequency range are reliably analyzed in both exome captures. 119 variants (figure below) were identified with p Conclusion These findings provide novel launch points for mechanistic studies that will further our understanding of disease pathogenesis. For example, SDF2L1 (R161H) flags a gene not previously implicated in IBD but which is reported to regulate feeding-induced ER stress, with a series of CRISPR screens identifying it as an essential regulator of ER homeostasis. Download : Download high-res image (141KB) Download : Download full-size image Effect size versus frequency for significant low frequency coding variants (credit: Hailiang Huang)