Chromosome Wide Prediction of Tissue Specific Enhancers Employing Computational and Statistical Approaches

Abstract

Eukaryotic genome is a regulatory jungle having various gene circuits or networks. Regulatory elements serve as switches in these gene circuits. Complete understanding of gene regulation can unravel the hidden mysteries of the genome. Seemingly junk or non coding portions of genome like gene deserts and interons can also hold crucial functional duty of gene regulation. Nevertheless, limited understanding of regulatory elements has hindered their identification always. However, various experimental and computational techniques have been developed to capture their presence indirectly if not directly. Cis-Regulatory Modules function in a highly tissue specific fashion. This tissue specificity is bestowed to CRMs by combinatorial nature of action of transcription factors. Transcription factors behave in combinatorial mode to achieve spatiotemporal specificity. Cis-Regulatory Modules bear binding sites for these transcription factors in same combinatorial way and inherit the spatiotemporal specificity too. We exploited same property of Cis-Regulatory Modules for their prediction. For this purpose we curated a catalogue of transcription factors using strong literature and experimental evidence. Taking into consideration the evolutionary relevance of human forebrain, we attempted to define forebrain specific transcription factor code. Co occupancy of these forebrain specific transcription factors with was used to locate forebrain specific Cis-Regulatory Modules in non-coding, non repetitive segments of chromosome 7 including gene deserts. We successfully predicted 1453 putative forebrain specific Cis-Regulaory Modules. On their comparison with experimentally verified enhancers from VISTA Enhancer browser, we found some of the enhancers overlapping with our putative CRMs. Developmental stage and conservation independence, low cost and time efficiency make our method of prediction a comparable substitute of already available computational and experimental techniques of Cis- Regulatory module detection.