Entering edit mode
As an example, if the study design is treatment, and there are 3 different groups of samples in the study, treatment A, treatment B, and treatment C. I know I can get pair-wise comparison by using results(dds, contrast=c("condition","Treatment A","Treatment B"), but I'm curious about how DESeq2 does that. Is DESeq2 constructing a linear model for each gene and each of the 3 conditions and comparing their coefficient pair-wise? Thanks.