We then report the variance explained; for quantitative traits, this is the variance explained by a linear model and for diseases, the variance explained is from a logistic model (Nagelkerke R2)2,3,64, which we convert to liability scale pseudo R2 such that R2 values are comparable among both quantitative and case–control phenotypes. We used various GWAS P value thresholds (0.1, 0.03, 0.01, 0.003, 0.001, 3 × 10−4, 1 × 10−4, 3 × 10−5, 1 × 10−5) to assess the predictive performance of our PRS.
