As expected, we observed across all scenarios that accuracies of PGS decreased monotonically with increased genetic distance to EUR (Supplementary Fig. 3). The genetic distance was measured as FST described in Supplementary Note 2. More specifically, we found the largest RA in individuals of Asian ancestry (mean RA ~91% in SAS and mean RA ~77% in EAS), which has an average FST of ~0.06 with EUR. The smallest RA was observed in individuals of AFR ancestry (~46%), which has an average FST of ~0.14 with EUR (Supplementary Figs. 3 and 4). These results imply that trans-ancestry predictive power of PGS remains limited even when causal variants and their effect sizes are shared between ancestries20,23. Across 18 simulation scenarios and over 100 simulation replicates for each scenario (Fig. 1), we found differences between the mean observed RA and the predicted RA from Eq. (1) to range between −4.5 and +2.5%. Although these differences were statistically significant in 11/18 simulation scenarios (two-tailed t-test, p-value <0.05/18), we found their sign to be inconsistent between ancestries. In fact, predicted RA from Eq. (1) slightly
