Evidence based housekeeping gene selection for microRNA-sequencing (miRNA-seq) studies

Abstract

The accurate determination of miRNA expression by PCR requires robust data normalisation. The most commonly used approach involves the use of housekeeping or normalising factors which are assumed invariant across the range of experimental conditions investigated. This assumption is often not proven in all experimental situations, therefore we consider whether global miRNA expression data sets obtained by high throughput sequencing can be screened to identify candidate housekeeping miRNAs (HK-miRNA) for further evaluation. To establish this method, we assessed global miRNA sequencing data from a study involving 30 canine skin specimens. From an initial pool of over 200 miRNAs, we identified several candidate HK-miRNA which demonstrated stability across all study samples using both a classical statistical approach and the NormFinder algorithm. We verified these putative HK-miRNAs using real-time PCR assays to further validate their suitability in this specific experimental setup. Our analysis provides a framework to allow researchers to exploit high-throughput sequencing data (be it their own, or from the short read archive or other genomic repository) to guide HK-miRNA selection.