The allele frequencies were compared between cases and controls in the European-American discovery sample using genome-wide logistic regression analysis as implemented in the program PLINK (Purcell et al., 2007). Diagnosis served as the dependent variable, alleles served as the independent variables, and ancestry proportions, study sites (COGA, COGEND and FSCD), sex, and age served as the covariates. The ancestry proportions for each individual were estimated by integrating the ancestry information content of 3,172 independent ancestry-informative SNPs (Zuo et al., 2012) using the program STRUCTURE. The Manhattan plot and the QQ plot were shown previously (Zuo et al., 2011, 2012). The most significant gene region (i.e., Na+/K+ transporting ATPase interacting 1 gene (NKAIN1) - serine incorporator 2 gene (SERINC2) region in the present study) was tested in the replication sample and other non-alcoholism samples, to examine its replicability and specificity (see below).
