Comprehensive analysis of RNA-chromatin, RNA- and DNA-protein interactions

RNA-chromatin interactome data is considered to be one of the noisiest types of data in biology. This is due to protein-coding RNA contacts and non-specific interactions between RNA and chromatin caused by protocol specifics. Therefore, finding regulatory interactions between certain transcripts and genome loci requires a wide range of filtering techniques to obtain significant results. Using data on pairwise interactions between these molecules, we propose a concept of triad interaction involving RNA, protein and a DNA locus. The constructed triads show significantly less noise contacts and are more significant when compared to a background model for generating pairwise interactions. RNA-chromatin contacts data can be used to validate the proposed triad object as positive (Red-ChIP experiment) or negative (RADICL-Seq NPM) controls. Our approach also filters RNA-chromatin contacts in chromatin regions associated with protein functions based on ChromHMM annotation.