We used the two GWAS summary statistics data from the largest PTSD GWAS study published to date conducted by PGC-PTSD:17 one performed on European ancestry cohorts and another on African ancestry. Each study consists of ~10,000 samples with ~25% PTSD positive cases. The African ancestry summary statistics did not result in any statistically significant predictive PRS, even when considering only African–American (or cluster 3) subset of our target subjects. This might arise from several technical challenges, including the fact that the African genome is highly diverse with short LD blocks and recent admixtures29–31, and most commercial arrays that tag a single variant from an LD block have low genetic coverage for African ancestry genome32. For this reason, only the European (EA)-based summary statistics are used in the present study to build PTSD-PRS for all target subjects including those of African ancestry participants.
