Methods). When we re-computed the risk scores using TWAS values for expression from the NTR cohort (which introduces additional noise due to heterogeneity between the cohorts), TWAS remained significantly higher than the eQTL strategy, but was comparable to selecting the nearest gene (Supplementary Table 5). This indicates that using expression from a different study to select genes still significantly explains variance, but is complementary (rather than superior) to selecting the nearest gene. Working from the assumption that genes with a higher cis-genetic correlation to phenotype are more likely to be causal, these results motivate the use of TWAS to prioritize putative risk genes at known GWAS loci.
