Chunk #61 — Methods — SNP and gene findings in the context of previous analyses — Overlap of previous rare coding variants in OCD and GWAS gene findings
First, we comprehensively assessed the overlap between 251 genes that we highlighted in our study as carrying common risk variation for OCD (Supplementary Table 14) and current gene-based summary statistics from OCD exome-sequencing data. We used data from Halvorsen et al.88 because it is the largest published exome-sequencing study of OCD presently. The supplementary materials from that paper include de novo variant calls from 771 case trios and 1,911 controls (supplementary table 14 in ref. 88). We compared the burden of de novo variants, partitioned by variant annotation (synonymous, missense, loss of function) in trio cases versus trio controls within these 251 GWAS-prioritized genes. As described previously88, we only included de novo variants that were in loci well covered in both case and control data (In_Jointly_Covered_Loci==TRUE). We also excluded all calls from quartet samples in ref. 88 (Cohort!=“OCD_JHU_quartets”). For each of the four variant annotation classes, we compared the proportion of cases that had at least one qualifying de novo variant to the proportion of controls using a two-sided Fisher’s exact test.
