For testing for a difference in the proportion of cell states in individuals with AUD, the mean proportion of cells in each cluster was calculated for each sample, and an ANOVA was performed to determine if the mean proportion significantly changed in samples from individuals with AUD as compared to those without. Age, sex, and ethnic origin were used as covariates. For this test, we removed samples with fewer than 50 cells of the cell type being tested. FDR correction of p values was performed using the BH procedure, as described in their original paper83, using an FDR cutoff of less than 0.05.
