Improving Reproducibility and Candidate Selection in Transcriptomics Using Meta-analysis

Abstract

Transcriptomic experiments are often used in neuroscience to identify candidate genes of interest for further study. However, the lists of genes identified from comparable transcriptomic studies often show limited overlap. One approach to addressing this issue of reproducibility is to combine data from multiple studies in the form of a meta-analysis. Here, we discuss recent work in the field of circadian biology, where transcriptomic meta-analyses have been used to improve candidate gene selection. With the increasing availability of microarray and RNA-Seq data due to deposition in public databases, combined with freely available tools and code, transcriptomic meta-analysis provides an ideal example of how open data can benefit neuroscience research.