Table 1. Motif predictions comparison.

	total TFs	True Positives (matched TFs)	Recall
ARACNe	70	27	38.57%
Coexpression	70	13	18.57%
Coexpression*	70	25	35.71%
ARACNe-MRC	70	25	35.71%

We compared motif enrichment accuracy across promoters corresponding to targets identified by the regulatory-network reverse-engineering algorithm ARACNe, co-expression, and a combination of both methods (see Results). ARACNe-MRC corresponding to ARACNe inferred target promoters with exons and repeats masked; see Table S1 for expanded description.