the trans-ancestry PRS, we compared the risk of T2D among individuals in progressively more extreme cutoffs of the PRS distribution relative to the rest of the samples. Individuals of European ancestry in the top decile of the trans-ancestry PRS had an OR of 3.19 (95% CI: 2.97–3.42; P value = 1.33E−232) compared with individuals in the bottom 90% of the PRS distribution. Risk continued to increase when contrasting the top 5% of the PRS (OR = 3.55, 95% CI: 3.24–3.90; P value = 1.88E−158) and the top 2% of the PRS (OR = 4.21, 95% CI: 3.66–4.84; P value = 1.82E−89) to the lower tails of the PRS distribution (Table 2; Additional File 2: Table S3). This shows that the trans-ancestry PRS can identify individuals of European ancestry with a significantly increased risk of T2D. Using the top 2% of the PRS as the classifier, the prevalence-adjusted PPV and NPV were 0.26 and 0.90, respectively (Table 2; Additional File 2: Table S3).Table 2Prediction accuracy of the trans-ancestry T2D PRS in eMERGE across three populationsPopulationLiabilityR2Covariates-adjusted AUCOR per SD(95% CI)Top 2% PRSOR(95% CI)P valueSensitivitySpecificityAdjusted PPV*Adjusted NPV*European9.2%0.661.96(1.91, 2.02)4.21(3.66, 4.84)1.82E−890.050.990.260.90African2.8%0.581.54(1.46, 1.64)1.98(1.43, 2.74)4.34E−050.030.980.210.88Hispanic8.0%0.632.08(1.84, 2.35)6.87(3.11, 15.15)1.81E−060.040.990.430.87*Adjusted PPV and NPV are calculated using the following population-specific prevalence:
