Accurate prediction of functional states of cis-regulatory modules reveals the common epigenetic rules in humans and mice

We proposed a two-step approach for predicting active cis-regulatory modules (CRMs) in a cell/tissue type. We first predict a map of CRM loci in the genome using all available transcription factor binding data in the organism, and then predict functional states of all the putative CRMs in any cell/tissue type using few epigenetic marks. We have recently developed a pipeline dePCRM2 for the first step, and now presented machine-learning methods for the second step. Our approach substantially outperforms existing methods. Our results suggest common epigenetic rules for defining functional states of CRMs in various cell/tissue types in humans and mice.