Classification of pre-cursor micrornas from different species using a new set of features

Abstract

MicroRNAs (miRNAs) are short RNA sequences actively involved in post-transcriptional gene regulation. Such miRNAs have been discovered in most eukaryotic organisms. They also seem to exist in viruses and perhaps in microbial pathogens to target the host. Drosha is the enzyme which first cleaves the pre-miRNA from the nascent pri-miRNA. Previously, we showed that it is possible to distinguish between pre-miRNAs of different species depending on their evolutionary distance using just k-mers. In this study, we introduce three new sets of features which are extracted from the precursor sequence and summarize the distance between k-mers. These new set of features, named inter k-mer distance, k-mer location distance and k-mer first-last distance, were compared to k-mer and all other published features describing a pre-miRNA. Classification at well above 80% (depending on the evolutionary distance) is possible with a combination of distance and regular k-mer features. The novel features specifically aid classification at closer evolutionary distances when compared to k-mers only. K-mer and k-mer distance features together lead to accurate classification for larger evolutionary distances such as Homo sapiens versus Brassicaceae (93% ACC). Including the novel distance features further increases the average accuracy since they are more effective for lower evolutionary distances. Secondary structure-based features were not effective in this study. We hope that this will fuel further analysis of miRNA evolution. Additionally, our approach provides another line of evidence when predicting pre-miRNAs and can be used to ensure that miRNAs detected in NGS samples are indeed not contaminations. In the future, we aim for automatic categorization of unknown hairpins into all species/clades available in miRBase.