In this study, we created a compendium of 707 cell-type-specific regulatory annotations to analyze 111 complex traits and diseases in EUR and EAS populations. We demonstrated that IMPACT annotations help pinpoint ancestrally portable genetic effects from association data. First, we showed that trait-associated annotations capture indistinguishable proportions of heritability between EUR and EAS populations. Second, we showed that these annotations implicate variants with higher trans-ancestry marginal effect size correlations, while negligibly affecting the distribution of Fst; this might explain the improvement driven by functional prioritization in P+T PRS models that use marginal effect sizes. Third, we showed that leveraging these annotations in PRS models improves accuracy, especially for the trans-ancestry application.
