As the scope of gene expression studies continues to expand, the need to quantify and interpret multiple drivers of expression variation is becoming essential. Here we present variancePartition, a publicly available software package that leverages the power of the linear mixed model to quantify the contribution of multiple sources of variation in complex gene expression datasets. For each gene, this analysis partitions the total expression variance into the fraction attributable to each aspect of the study design. A variancePartition analysis gives a genome-wide summary of the drivers of variation, but also produces gene-level results to identify genes that deviate from the genome-wide trend.
