Accurate diagnosis of cases and controls is crucial for screening PGS. Therefore, we used 3 datasets from cohorts ascertained to study AUD (COGA, SAGE, and OZALC). AUD was determined using Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) criteria in the COGA cohort and DSM, Fourth Edition (DSM-IV) criteria in the SAGE and OZALC cohorts. Those who did not meet any DSM-IV or DSM-5 criteria were considered controls. Additionally, because AUD is substantially genetically correlated with other substance use disorders, we also excluded those with other substance use disorders from the controls. These 3 datasets were combined as the screening dataset.
