Figure 2

PEAKS [29]output patterns. (A) 3,061 consistent nucleosomes with > 95% overlap between two independent experimental datasets [7, 8]; (B) 3,288 weakly positioned nucleosomes with < 5% overlap. Both sets were centrally aligned around the position of the assumed pseudodyad axis, representing 0 in the x axis. Both forward and reverse complements of each sequence were included in the analysis. Colored bubbles correspond to positional biases of significant regulatory motifs as defined by Harbison et al. [30]. Each significant motif is represented as an oval, colored according to an arbitrary internal legend. The width of each oval corresponds to the range of the position of significance, and the height is the relative motif signal (RMS). The RMS is defined as the maximum number of sequences that contain the motif in the region over the number of sequences that contain the motif at the P-value cut-off level; that is, the random expectance of the motif occurring in that region in 1000 random sets (for example, the elongated red oval for Abf1 in (A) was found to a significant bias in positions 120 to 150 nucleotides away from the center of the nucleosome, and in 2.5 times more sequences than those expected under a random model, thus having an RMS value of 2.5). Notice the difference in the scale between the two panels, suggesting increased fold biases for the consistent nucleosomes. Images produced by PEAKS are reproduced here with the permission of the program's authors.