Differential gene expression analysis between samples from individuals with and without AUD was performed for each cell type, as well as the two subclusters of astrocytes and four subclusters of microglia, using DESeq2,98 a statistical package designed for bulk RNA-seq data, with the default parameters. Briefly, the tool estimates the variance of gene expression and then fits a negative binomial distribution to each gene, which accounts for the over-dispersion of RNA sequencing data, which can result in more accurate p-values. Sex, age (as a continuous variable), and ethnic origin were included as covariates in the models. Genes with p values of less than 0.2 (corrected for multiple-hypothesis testing using the Benjamini-Hochberg method) were deemed significant.
