We used cross-trait LDSC31, a method that computes genetic correlations between GWASs without bias from ancestry differences or sample overlap to calculate genetic correlations between the primary OCD meta-analysis and other phenotypes of interest. The selection of traits was based on phenotypic relevance and/or prior report of a genetic relationship with OCD. The genetic correlation between traits was based on the estimated slope from the regression of the product of Z scores from two GWASs on the linkage disequilibrium score and represents the genetic covariation between two traits based on all polygenic effects captured by the included SNPs. The genome-wide linkage disequilibrium information used by these methods was based on European populations from the HapMap 3 reference panel101, and GWAS summary statistics were filtered to only include SNPs that were part of the 1,290,028 HapMap 3 SNPs.
