Distribution of significance score values returned by dyad-analysis in promoters of orthologous clusters (solid lines) or in random selections of genes (dotted lines). Each curve represents the results of the analysis at one taxonomical level. A. Number of dyads per sequence set. B. Number of sequence sets with at least one significant motif. Black lines represent the theoretic curves. We run dyad-analysis with the TAXFREQ background model and the E. coli-specific dyads filtering.