Fig. 4

The most prevalent contextual associations of EVE proteins in the genomes of Proteobacteria. A contextual information network graph of domains detected among genes in the genomic neighborhoods encoding EVE proteins in Proteobacteria. The nodes are clustered by mutual connections, which generally correlate with the distribution of species in which they occur. The thickness of the edges reflects the strength of the association. Nodes representing domains that participate in the same pathway are drawn with the same color. Domain coloring and abbreviations are explained in the text and Figs. 3 and 5. The graph was calculated from the top 400 pairwise associations between domains in EVE protein neighborhoods from 13,388 genomes with unique domain compositions. Singleton associations with EVE and minor networks were removed for clarity