Skip to main content
Figure 4 | BMC Bioinformatics

Figure 4

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

Figure 4

Effect of increasing surrounding sequence length on identification of the five known skeletal muscle TFBSs. TFBS motifs identified using: (A) the sequence conserved between the human and mouse genomes within the human CRM sequence sets previously described by Wasserman et al. [15]; (B) the sequence conserved between the human and mouse genomes within the first 1000 bp upstream of the same genes for which CRMs were analyzed in (A); (C) as in (B), except examining 2000 bp upstream of the same gene set; (D) as in (B), except examining 5000 bp upstream of the same gene set. For each of the five TFs, the logo of the previously described motif [35] is shown, alongside the logo of the highest ranking (according to group specificity) motif matching the given TF, the discovered motif's rank and its group specificity score. Five motif finding runs were performed for each of these four input sequence sets; if a given motif was discovered multiple times over the five runs, then the ranks of both the collapsed motifs and also the highest rank from the individual runs are shown, along with the geometric mean of the group specificity scores from individual runs in which the motif was found and the logo of the best scoring motif from the individual runs. Shown below each set of discovered motifs is the geometric mean and range of the group specificity scores of motifs discovered within five size-matched sets of randomly selected regions. All results shown here are for motifs that passed our block filtering criteria.

Back to article page