Improved moderation for gene-wise variance estimation in RNA-Seq via the exploitation of external information

Abstract

Background: The cost of RNA-Seq has been decreasing over the last few years. Despite this, experiments with fouror less biological replicates are still quite common. Estimating the variances of gene expression estimates becomesboth a challenging and interesting problem in these situations of low replication. However, with the wealth ofmicroarray and other publicly available gene expression data readily accessible on public repositories, these sourcesof information can be leveraged to make improvements in variance estimation.Results: We have proposed a novel approach called Tshrink+ for inferring differential gene expression throughimproved modelling of the gene-wise variances. Existing methods share information between genes of similaraverage expression by shrinking, or moderating, the gene-wise variances to a fitted common variance. We havebeen able to achieve improved estimation of the common variance by using gene-wise sample variances fromexternal experiments, as well as gene length.Conclusions: Using biological data we show that utilising additional external information can improve themodelling of the common variance and hence the calling of differentially expressed genes. These sources ofadditional information include gene length and gene-wise sample variances from other RNA-Seq and microarraydatasets, of both related and seemingly unrelated tissue types. The results of this are promising, with ourdifferential expression test, Tshrink+, performing favourably when compared to existing methods such as DESeqand edgeR when considering both gene ranking and sensitivity. These improved variance models could easily beimplemented in both DESeq and edgeR and highlight the need for a database that offers a profile of genevariances over a range of tissue types and organisms.