Figure 1.

ReMoDiscovery analysis flow. ReMoDiscovery consists of a seed discovery step followed by a seed extension step. ChIP-chip data, motif data, and expression data are used as input for the algorithm. These three datasets can be represented as matrices in which the rows represent the genes. For the ChIP-chip data (R) the columns represent the regulators, for the motif data (M) they represent the motifs and for the expression data (A) the different experiments. (a) The seed discovery step identifies sets of genes that are co-expressed, bind the same regulators, and have the same motifs in their intergenic region. (b) The gene content of the seed modules can be extended during the seed extension step using less stringent criteria. The logarithms of the module enrichment p values (y-axis) are plotted for all regulators (motifs) as a function of the correlation threshold (x-axis). Each line in the sample plot shows the module enrichment p values for the enrichment of its corresponding regulator (motif) as a function of the gene expression correlation threshold used.