Reconstruction of novel transcription factor regulons through inference of their binding sites

In most sequenced organisms the number of known regulatory genes (e.g., transcription factors (TFs)) vastly exceeds the number of experimentally- verified regulons which could be associated with them. At present, identification of TF regulons is mostly done through comparative genomics approaches. The nature of such methods causes them to frequently miss organism-specific regulatory interactions and often requires expensive and time-consuming experimental techniques to generate the underlying data. One approach to computationally addressing these problems is though discovery of transcription factor binding sites (TFBSs) and inference of corresponding regulons based on the location of such motifs across the genome. In this work, we present an efficient algorithm that aims to identify a given transcription factor's regulon through inference of its unknown binding sites, based on the discovery of its binding motif, which is also unknown and is estimated within the algorithm framework. The proposed approach relies on computational methods that utilize microarray gene expression data sets and fitness data sets which are available or may be straightforwardly obtained for many organisms. We computationally constructed the profiles of putative regulons for the TFs LexA and PurR in E. coli K12 and identified their binding motifs. Comparisons with an experimentally-verified database showed high recovery rates of the known regulon members (93% recovery for LexA regulon), and indicated good predictions for the newly found regulon genes with high biological significances. The results also show that the algorithm can predict genome-wide transcription factor binding motifs, which display high homology to their known consensus binding sites. The proposed approach is also applicable to novel organisms for predicting unknown regulons of the transcriptional regulators.