Update (21/02/2014) The voom method has now been published in Genome Biology
Original post Gordon Smyth is well known for his development of the limma Bioconductor package for the analysis of differential gene expression using microarrays. More recently his group has led the way in the development of software for the statistical analysis of gene expression using RNA-seq with the edgeR Bioconductor package and the voom() method in limma. Today, Gordon spoke about modeling the variance in RNA-seq data for studying gene expression, in particular using the voom() method, and contrasted this approach with that taken by edgeR (and other Poisson/Negative binomial-like methods).