models using generalized estimating equations; specifically, we included a random intercept to adjust for the family relationships. Unrelated samples from the AOU and IB datasets were included; thus, we used logistic regression. Sex, the first 10 principal components of genetic ancestries, and genotyping arrays in each dataset were included as covariates. These datasets used different definitions of age; thus, age was not included as a covariate, as in a previous large AUD GWAS.45 SAS, version 9.4 (SAS Institute Inc), was used to perform all statistical analyses. Additionally, for the best PGS, we tested estimabilities in male and female individuals separately. Because the rs1229984 T allele in ADH1B is the largest protective genetic factor for AUD in populations of European ancestry,46 we also tested the best PGS by excluding carriers of the rs1229984 T allele.
