The model.

<p>(a) The mechanism of interaction between the genes as envisaged in our model. The genes are indicated by ellipses (green if they code transcription factors (TFs), blue otherwise), the TFs by triangles with the associated binding motif (regulatory sequence) in the box underneath. Non-TF proteins are symbolized by the “P” shape, and the promoter regions (PRs) upstream of each gene are shown as red boxes. Binding occurs if the binding motif exactly matches a subsequence in the PR, as is the case here at PR4. PRs in the model are typically much longer than depicted here. (b) Distribution of the amount of bit-wise information coded by each regulatory sequence recognized and bound by the 102 TFs in the yeast genome, compiled from the recently published data by Harbison et al. <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0000501#pone.0000501-Harbison1" target="_blank">[9]</a>. This distribution is adopted as the length distribution of the random regulatory sequences (“binding motifs”) in our model.</p>