Table 1 Comparison of different regions of sequence conservation in the muscle related genes. (A) Number of experimentally verified TFBSs found in the conserved regions of different sizes. The Wasserman conserved regions were taken from Wasserman et al. [15]; note that the regions were identified using a Bayesian alignment method between human and mouse, while the upstream regions that we call conserved are from the UCSC Bioinformatics Site [29,30]. The 1, 2, 5 kb conserved regions were generated using the blastz alignment of the human (hg16) and mouse (mm4) taken from the UCSC Bioinformatics Site. The name of each TF family is listed next to the number of known sites taken from Wasserman et al. [14]. The number of sites found within the CRMs is listed next. The number of sites found in the sequences generated by SequenceExtractor is listed according to the length of the region. (B) Number of sites matching the corresponding TFBS motif, at a Pearson correlation coefficient threshold of 0.6, within each of the regions shown in (A). (C) Relative enrichment of each TFBS motif in the specified regions, normalized to the CRMS. Specifically, the enrichment is calculated by dividing the frequency of the motif in the region of interest by its frequency in the CRMs.

From: Meta-analysis discovery of tissue-specific DNA sequence motifs from mammalian gene expression data