Ethanol treatment was analyzed as paired samples (treated and untreated samples from the same individual), which eliminated the need to use gender, phenotype and batch as factors. False discovery rates (FDR) were calculated using the Benjamini & Hochberg (Benjamini & Hochberg, 1995) method within edgeR. EdgeR paired analysis does not do a paired interaction of group (AD vs. control) × exposure, therefore we calculated the CPM treated – CPM untreated for each individual for each gene and used a t-test to see if the mean difference for each gene was equal between alcoholics and controls. This same method was used to test for gender differences in response to alcohol.
