10.1101/856898

Accurate modeling of replication rates in genome-wide association studies by accounting for winner's curse and study-specific heterogeneity