A major advantage of our study is having large simulated European and African ancestry cohorts to provide guidelines for developing the best possible PRSs in admixed individuals with current and future GWASs. Through our exploration of 12 PRSs, with various variant selection and weighting approaches, we recapitulate recent results applying similar PRS strategies to an admixed Hispanic/Latino population.9 For individuals with intermediate proportions of European ancestry (20%–80%), we also see improvements using European selected variants and population-specific or fixed-effects meta-analysis weights; however, as non-European cohorts get increasingly large, it will be imperative to perform variant discovery in these populations, as gains in accuracy with weight adjustment of European selected variants will be limited, especially in individuals with higher proportions of non-European ancestry.
