Skip to main content


Figure 2 | BMC Bioinformatics

Figure 2

From: Predicting tissue specific cis-regulatory modules in the human genome using pairs of co-occurring motifs

Figure 2

Performances of CrmMiner and CisModule on controlled data sets. The performance is measured in terms of the true positive count, the false positive count, and the precision (Equation 9). A set name indicates the number of folds of background sequences mixed with the training and validation hypersensitive sites (HSSs). For instance, the training mixed set of "x10" included 740 HSSs located in the vicinity of 75 genes specific to the human CD4+ T cells. In addition to the 740 HSSs, this training mixed set contained 750 (10 × 75) background sequences. All testing mixed sets consisted of 443 HSSs i.e. they were not mixed with background sequences. All testing control sets were composed of 1125 random genomic sequences. The exact number of CRMs among the HSSs is unknown; however, HSSs are enriched with CRMs.

Back to article page