The MVP PTSD PRS did not predict PTSD diagnosis among either EA or AA COGA sub-samples. It is possible that this is a result of different sample composition for the MVP and COGA. For example, the MVP sample is predominantly male and consists of military veterans [22], whereas the COGA sample has closer to equal numbers of male and female participants and is a civilian population. While the MVP is a large and ancestrally diverse GWAS, it is possible that the relatively smaller AA samples in both the MVP and COGA study, could affect the MVP PTSD-PRS associations in the COGA sample. Clearly, more ancestrally diverse GWAS are needed to understand and replicate these findings.
