EdgeR employs statistical models to estimate the variability in gene expression counts, allowing for the identification of differentially expressed genes. It uses negative binomial distribution to model the count data, which is ideal for RNA-Seq data that often shows overdispersion.
