EPIGENE: genome wide transcription unit annotation using a multivariate probabilistic model of histone modifications

BackgroundUnderstanding transcriptome is critical for explaining functional as well as regulatory roles of genomic regions. Current methods for the identification of transcription unit (TU) uses RNA-seq which, however, requires large quantities of mRNA limiting the identification of inherently unstable TUs e.g. for miRNA precursors. This problem can be resolved by chromatin based approaches due to a correlation between histone modifications and transcription. ResultsHere we introduce EPIGENE, a novel chromatin segmentation method for the identification of active TUs using transcription associated histone modifications. Unlike existing chromatin segmentation approaches, EPIGENE uses a constrained, semi-supervised multivariate hidden markov model (HMM) that models the observed combination of histone modifications using a product of independent Bernoulli random variables, to identify active TUs. Our results show that EPIGENE can identify genome-wide TUs unbiasedly. EPIGENE predicted TUs showed an enrichment of RNA Polymerase II in transcription start site and gene body indicating that they have been transcribed. Comprehensive validation with existing annotations revealed that 93% of EPIGENE TUs can be explained by existing gene annotations and 5% of EPIGENE TUs in HepG2 can be explained by microRNA annotations. EPIGENE outperforms existing RNA-Seq based approaches in TU prediction precision across human cell lines. Finally, we identify 381 novel TUs in K562 and 43 novel cell-specific TUs all of which are supported by RNA Polymerase II data. ConclusionsWe demonstrate the applicability of HMM to identify genome-wide active TUs and provides valuable information about unannotated TUs. EPIGENE is an open-source method and is freely available at: https://github.com/imbeLab/EPIGENE.