We have developed a Bioconductor package edgeR that addresses one of the fundamental downstream data analysis tasks for count-based expression data: determining differential expression. The package and methods are general, and can work on other sources of count data, such as barcoding experiments and peptide counts. To the authors' knowledge, edgeR is the only software for SAGE or DGE data at this time which can account for biological variability when there are only one or two replicate samples.
