When studying the transcriptome, most of our inferences revolve around changes in average expression. However, more recent examples have demonstrated that analysis of the variability of gene expression can also highlight important regulators too. In this talk, I outline some of the bioinformatics methods my lab has developed to investigate the functional consequences of gene expression variability to understand transcriptional regulation. I present a recently published method called pathVar, which provides functional interpretation of variability changes at the level of pathways and gene sets. Application of pathVar to cancer patient cohort data will be shown to demonstrate the utility of this method. I also describe a method based on the third statistical moment, skewness, to model heterogeneously expressed genes. Using skewness-based metrics, we can uncover new genes with regulatory roles in cancer, as well as those that vary with DNA methylated loci. Collectively, this series of related studies outline the value