Computational identification of transcription factor binding sites by functional analysis of sets of genes sharing overrepresented upstream motifs

摘要

BACKGROUND: Transcriptional regulation is a key mechanism in the functioningof the cell, and is mostly effected through transcription factors binding tospecific recognition motifs located upstream of the coding region of theregulated gene. The computational identification of such motifs is made easierby the fact that they often appear several times in the upstream region of theregulated genes, so that the number of occurrences of relevant motifs is oftensignificantly larger than expected by pure chance. RESULTS: To exploit thisfact, we construct sets of genes characterized by the statisticaloverrepresentation of a certain motif in their upstream regions. Then we studythe functional characterization of these sets by analyzing their annotation toGene Ontology terms. For the sets showing a statistically significant specificfunctional characterization, we conjecture that the upstream motifcharacterizing the set is a binding site for a transcription factor involved inthe regulation of the genes in the set. CONCLUSIONS: The method we propose isable to identify many known binding sites in S. cerevisiae and new candidatetargets of regulation by known transcription factors. Its application to lesswell studied organisms is likely to be valuable in the exploration of theirregulatory interaction network.