Figure 7

Non-responsive gene promoters rarely contain Hap2/3/4/5p regulatory elements and respond to signals that depend on growth rate. (a) Frequency distribution of potential Hap2/3/4/5p sites in the promoters of genes from the responsive gene cluster (red) and the non-responsive gene cluster (blue) from Figure 6. The frequency of occurrence of potential Hap2/3/4/5p sites as detected by a Hap2/3/4/5/p nucleotide-distribution matrix was calculated in 100 bp intervals (threshold = 0.85 [24]). The distribution of Hap2/3/4/5p sites in the promoters of genes from the responsive gene cluster localize between -400 and -100 bp from the ATG, a common distance for yeast UAS. Putative Hap2/3/4/5p sites extracted from the non-responsive gene cluster show a random distribution similar to that found for random-sequence DNA (A/T 31%, C/G 19%), and are therefore likely to be non-functional matches to the matrix. (b) The correlation between Z-scores and growth rate of the constitutive (left) and regulated (right) clusters calculated with the Quontology algorithm (see also Figure 5) using experimental data from a large number of deletion mutants [15]. Z-scores larger than 4 or smaller than -4 have significantly altered expression (p < 0.1). The running mean of the absolute Z-scores was calculated to determine to what extent regulation is correlated to the growth rate.