EdgeR is a Bioconductor package for analyzing differential expression in digital gene expression (DGE) data. It uses an overdispersed Poisson model to account for both biological and technical variability, and employs empirical Bayes methods to moderate overdispersion across transcripts, improving reliability. The package can handle minimal replication, provided at least one group is replicated. It is applicable beyond sequencing data, such as proteome peptide counts and barcoding experiments. The software is freely available under the LGPL license. It is designed for detecting changes between groups using exact tests adapted for overdispersed data. The model assumes negative binomial distribution for count data, with parameters including library size, dispersion, and relative abundance. EdgeR estimates gene-wise dispersions using conditional maximum likelihood and moderates them using empirical Bayes. It supports pairwise comparisons and can be extended to multiple groups. For technical replicates, Poisson-based analysis is possible. EdgeR is the only software for SAGE or DGE data that accounts for biological variability with minimal replicates. Funding and conflict of interest statements are provided.EdgeR is a Bioconductor package for analyzing differential expression in digital gene expression (DGE) data. It uses an overdispersed Poisson model to account for both biological and technical variability, and employs empirical Bayes methods to moderate overdispersion across transcripts, improving reliability. The package can handle minimal replication, provided at least one group is replicated. It is applicable beyond sequencing data, such as proteome peptide counts and barcoding experiments. The software is freely available under the LGPL license. It is designed for detecting changes between groups using exact tests adapted for overdispersed data. The model assumes negative binomial distribution for count data, with parameters including library size, dispersion, and relative abundance. EdgeR estimates gene-wise dispersions using conditional maximum likelihood and moderates them using empirical Bayes. It supports pairwise comparisons and can be extended to multiple groups. For technical replicates, Poisson-based analysis is possible. EdgeR is the only software for SAGE or DGE data that accounts for biological variability with minimal replicates. Funding and conflict of interest statements are provided.