microTSS: Accurate microRNA transcription start site identification reveals a significant number of divergent pri-miRNAs
Author
Georgakilas, G.; Vlachos, I. S.; Paraskevopoulou, M. D.; Yang, P.; Zhang, Y.; Economides, A. N.; Hatzigeorgiou, A. G.Date
2014Keyword
Abstract
A large fraction of microRNAs (miRNAs) are derived from intergenic non-coding loci and the identification of their promoters remains 'elusive'. Here, we present microTSS, a machine-learning algorithm that provides highly accurate, single-nucleotide resolution predictions for intergenic miRNA transcription start sites (TSSs). MicroTSS integrates high-resolution RNA-sequencing data with active transcription marks derived from chromatin immunoprecipitation and DNase-sequencing to enable the characterization of tissue-specific promoters. MicroTSS is validated with a specifically designed Drosha-null/conditional-null mouse model, generated using the conditional by inversion (COIN) methodology. Analyses of global run-on sequencing data revealed numerous pri-miRNAs in human and mouse either originating from divergent transcription at promoters of active genes or partially overlapping with annotated long non-coding RNAs. MicroTSS is readily applicable to any cell or tissue samples and constitutes the missing part towards integrating the regulation of miRNA transcription into the modelling of tissue-specific regulatory networks. © 2014 Macmillan Publishers Limited. All rights reserved.