Figure 2.

Illustration of the set-cover based algorithm for the identification of one-vs-all discriminative subnetworks. Here, the coverage provided by a hypothetical three-gene subnetwork is shown. In the left panel, the distribution of the expression levels of each gene across all samples is shown. We first compute the mean (μi) and the standard deviation (σi) of this distribution for each gene gi. Subsequently, we identify samples that are positively or negatively covered by each gene. A gene gi with expression greater than μi + α * σi in a sample is said to positively cover that sample, while a gene gi with expression less than μi - α * σi in a sample is said to negatively cover that sample (we set α = 2 in our experiments). The negative and positive cover sets for each gene and the subnetwork composed of g1, g2 and g3 are shown on the right. In this example, this subnetwork negatively covers (all samples in) Stage B.

Erten et al. BMC Proceedings 2012 6(Suppl 7):S1   doi:10.1186/1753-6561-6-S7-S1