Figure 2.

Schematic representation of the motif prediction procedure employed in this study. For each species TUs were predicted and used for a COT prediction. The upstream regions of the COTs were analyzed using MEME. Following MEME analysis, predicted regulatory motifs were compared using COMPASS. The upstream regions containing significantly similar motifs were re-analyzed by MEME. This procedure was iterated until all identified motifs could be considered unique. The unique motifs of both sets were compared and on basis of this comparison, the motifs were divided into three different classes. All motifs were validated using MAST against other genomic sequences. Finally, regulons were predicted by scanning the genome with the identified motifs.