Human microRNAs exhibit two distributions of expression within and across cancer samples

It has been previously established that mRNA expression levels can be separated into two distinct classes. We explore whether the same pattern is observed in microRNAs. We used microRNA-Seq data from The Cancer Genome Atlas to examine the overall profile of microRNA expression data. Our results indicate that microRNA expression can be separated into two distributions but that these distributions are quite different from those of messenger RNAs. In addition, different microRNA expression classes seem to have differing abilities to distinguish tumors within and across cancer types. These trends are consistent across the entirety of the TCGA datasets in our analysis.