InSilico-ChIP: A Coregulation and Evolutionary Conservation Based Transcription Factor and Target Gene Predictor

Abstract

Chromatin Immunoprecipitation followed by high-content sequencing (ChIP-seq) is a powerful approach for identifying bonafide transcription factor (TF) binding sites, however these studies can be difficult, time-consuming and costly. They require ChIP-validated antibodies and a priori knowledge of which TF to pull down. Moreover, to gain further mechanistic insights, transcriptomic data is required to determine if TF binding alters proximal gene expression. Determining regulatory pathways from expression data is not an easy task. The typical result of a gene expression experiment, a set of co-regulated genes, may not include the relevant TF at all. Many such factors become active through mechanisms other than a change in their level of expression. To understand how such a set of genes is co-regulated, it is necessary to find evidence of shared TF binding sites (TFBS). This approach presents its own set of problems. TFBS are identified by short sequences that can exist by chance without being biologically functional. An evolutionary perspective is required to consider only those functionally important sites that are conserved between the promoters of the genes in question and those of their orthologs in related species. Furthermore, the common occurrence of short TFBS makes it necessary to consider only those TFs that are significantly more common in the co-regulated genes than in the genome (or microarray) as a whole. InSilico-ChIP is a precomputed database of evolutionarily conserved TFBS for various species, which accepts a set of genes and quickly returns the conserved TFs that are statistically over-represented in the proximal promoter regions of those genes. It allows new species data to be created using only a whole genome alignment with a related species and gene locations. Several methods of identifying TF binding sites can be used, varying in alignment type, location, conservation restrictions, and TFBS matrices used to analyze the promoter regions.