Our study has several limitations: (i) the gap in the prediction accuracy between European/Asian and African populations remains considerable, likely due to the underrepresentation of African samples in the training GWAS; (ii) evaluation samples other than Europeans, Africans and East Asians were limited; (iii) the selection of high-risk individuals at extreme tails of the PRS distribution may be associated with uncertainties [49], which has not been fully characterized in this work; (iv) the sample characteristics of the evaluation cohorts may not fully represent the patient group to which PRS results will be returned, and the effects of PRS may depend on ascertainment and vary across different target samples; (v) predictive performance of the PRS was only assessed for prevalent T2D cases; and (vi) the associations between the trans-ancestry PRS and standard clinical risk factors (e.g., BMI, low physical activity, and hypertension), and the value of the PRS over and above these factors remain unclear. Evaluating the capability of the trans-ancestry PRS in identifying incident cases (i.e., individuals at risk to develop the disease in a future time window) in prospective
