To find genes likely to contribute to AUD risk, we performed an integrative analysis by combining the differentially expressed gene list with GWAS findings and cell type-specific eQTL loci. Our assumption is that if a genetic variant is associated with the expression levels of a nearby gene (eQTL) and the gene locates in a GWAS locus of an AUD-related trait, the increased or decreased expression level of the gene is more likely to contribute to the trait. Several large-scale GWAS have found genetic loci associated with AUD-related traits: 496 independent loci associated with number of drinks per week,3 and 90 independent loci associated with PAU5, including 5 loci associated with both traits. There are 3,406 and 749 genes, respectively, within these loci, of which 147 were associated with both traits, a total of 4008 unique genes (Supplementary Table 13). Of these, 518 were differentially expressed (padj < 0.2) in astrocytes in our snRNA-seq data, 861 in oligodendrocytes, 318 in D1 neurons, and 329 in D2 neurons. Differentially expressed genes within an associated locus are more likely than the others to
