rVISTA 2.0: evolutionary analysis of transcription factor binding sites

Abstract

Identifying and characterizing the transcription factor binding site (TFBS) patterns of cis-regulatory elements represents a challenge, but holds promise to reveal the regulatory language the genome uses to dictate transcriptional dynamics. Several studies have demonstrated that regulatory modules are under positive selection and, therefore, are often conserved between related species. Using this evolutionary principle, we have created a comparative tool, rVISTA, for analyzing the regulatory potential of noncoding sequences. Our ability to experimentally identify functional noncoding sequences is extremely limited, therefore, rVISTA attempts to fill this great gap in genomic analysis by offering a powerful approach for eliminating TFBSs least likely to be biologically relevant. The rVISTA tool combines TFBS predictions, sequence comparisons and cluster analysis to identify noncoding DNA regions that are evolutionarily conserved and present in a specific configuration within genomic sequences. Here, we present the newly developed version 2.0 of the rVISTA tool, which can process alignments generated by both the zPicture and blastz alignment programs or use pre-computed pairwise alignments of several vertebrate genomes available from the ECR Browser and GALA database. The rVISTA web server is closely interconnected with the TRANSFAC database, allowing users to either search for matrices present in the TRANSFAC library collection or search for user-defined consensus sequences. The rVISTA tool is publicly available at http://rvista.dcode.org/.

Figure 1

rVISTA 2.0 analysis data flow. (A) The rVISTA tool can process blastz alignments submitted at the rVISTA homepage (http://rvista.dcode.org/), or alignments automatically forwarded from the zPicture alignment program (http://zpicture.dcode.org/), the ECR Browser (http://ecrbrowser.dcode.org/) or the GALA database (http://globin.cse.psu.edu/gala/). (B) Users select the search criteria, and the results are returned in the same page as the downloadable static data files and dynamic links to visual analysis of TFBS distribution. (C) TFBSs for pre-selected TFs can be visualized above the conservation profile as tick marks, and the clustering module can detect user-specified groups of TFBSs.

Figure 2

TFBS analysis of the NKX2.5 genomic locus. (A) The NKX2.5 genomic region was accessed in the ECR Browser. Human/mouse and human/rat alignments are displayed (7 kb in the window). (B) Coding exons are in blue, untranslated regions (UTRs) are in yellow, conserved intronic noncoding ECRs are in pink and conserved intergenic ECRs are in red. The alignment was processed for Smad4 binding sites. (C) Smad4 TFBS matches to the reference sequence (human) are in blue, aligned pairs in red and aligned-and-conserved in green. NKX2.5 cardiac enhancer harbors 4 conserved Smad4 sites; one site corresponds with a previously functionally characterized Smad4 site.