Parameter landscape analysis for:


Identifying regulatory networks by combinatorial analysis of promoter elements

Pilpel, Sudarsanam & Church


The following Supplementary figures summarize analyses of the effect of various
thresholds and settings

Using ScanACE we assigned each motif in the collection to yeast promoters
Promoters scoring higher than 2 std below the mean ScanACE scores of the set of genes
from which the motif were derived constitute the original motif-promoter assignment in the paper.
We have then set various percentage thresholds on the ScanACE scores in order to test the effect
of considering promoters with higher and higher ScanACE scores

In Supllementary Figs 1-3 we analyse the impact of thresholds on different outcomes of
the analysis.
Supllementary Fig1 The expression coherence scores is calculated for each partner in each of 115
synergistic motif pairs at various threshold of the ScanACE cut-offs
Supllementary Fig2 The synergy scores calculated for each of the 115 synergistic motif pairs
115 at the same thresholds on the ScanACE cut-offs as in Supllementary Figure 1
Supllementary Fig3 A set of distributions of the synergy scores calculated for each of 115 synergistic
pairs at the same ScanACE thresholds as in figs 1 and 2 above.
Supllementary table1 The effect of modiying the threshold on distance between "close" ORF profile pairs. The original definition was - a pair of profiles is considered "close" if the Eicledean distance between their profiles is lower than the 5th percentile distance obtained for two random ORFs. The figure here shows the effect of synergy scores of using three other threshold - of 10, and 15 percentiles. Each line corresponds to one of the 115 synergistic pairs. The columns corresponds (from left to right) to threshold 5 (the original) 10, and 15 percentiles