Figure 9
From: MTGO: PPI Network Analysis Via Topological and Functional Module Identification
Iteration Phase of MTGO. Nodes are assigned to topological modules c h (Panel A). Functional modules δ fit topological modules differently. For example, δ1,1, δ2,1, and δ3,1, overlap differently with c1. The best functional module is δ3,1, since it minimizes the number of nodes out of the intersection between c1 and itself. It is then selected as δB,1 (Panels B and C). Once δB,1 is selected, the nodes of δB,1∪c1 are grouped into three sets: V a , V b , and V c (Panel D). V a are the nodes shared by δB,1 and c1; V b are the nodes belonging to c1 but not to δB,1; V c are the nodes belonging to δB,1 but not to c1. V a nodes stay in c1; V b nodes are moved to the TNL; V c nodes either remain in their topological module c3, or are moved to c1, according to the Modularity Variation function. Here, one V c node is embedded in c1, while the other stay within its original topological module c3.
