Computational-based approaches in epigenetic research: Insights from computational tools, mathematical models, and machine learning methods

Understudying epigenetics underlying mechanisms is essential. Studies have indicated functional roles of epigenetic events, including DNA methylation and histone modifications, in human disease, stem cell growth and development, aging, response to environmental stresses, and species evolution. High-throughput sequencing techniques alongside routine experimental approaches have rapidly produced a bulk of data that the major part of them remained unprocessed. To analyze, interpret, and process of these data, availability of efficient computational methods is critical. Epigenetic data analysis is complex and difficult because these data contain a multi-layer set of information. The methods implemented for this purpose must be able to handle massive data of experimental works and process the epigenetic layers. Furthermore, the methods must be capable of integrating multiple modifications and their combination effects on chromatin conformational structure and consequently the expression network of genes. In this study, we briefly reviewed challenges in the way of the computational epigenetics, the latest reported methods, and significant biological results derived from applying computational-based methods on epigenetic data.

A Short Report on the Markov Property of DNA Sequences on 200-bp Genomic Units of ENCODE/Broad ChromHMM Annotations: A Computational Perspective

The non-coding DNA in eukaryotic genomes encodes a language which programs chromatin accessibility, transcription factor binding, and various other activities. The objective of this short report was to determine the impact of primary DNA sequence on the epigenomic landscape across 200-base pair genomic units by integrating nine publicly available ChromHMM Browser Extensible Data files of the Encyclopedia of DNA Elements (ENCODE) project. The nucleotide frequency profiles of nine chromatin annotations with the units of 200 bp were analyzed and integrative Markov chains were built to detect the Markov properties of the DNA sequences in some of the active chromatin states of different ChromHMM regions. Our aim was to identify the possible relationship between DNA sequences and the newly built chromatin states based on the integrated ChromHMM datasets of different cells and tissue types.