ABSTRACT
Often when viewing a 2-D or 3-D plot of a complex system’s graph, the eye can pick out clusters of nodes that apparently “talk more amongst themselves” than to other peripheral nodes. The local density of the edges is the give-away in this subjective identiŸcation of astructural module, or subsystem. Modularity is an abstract concept that can refer to patterns of physiological and molecular interactions, the organization of graphs for modeling living processes, or in genetics to the distribution of mutational effects on the phenotype (Wagner et al. 2007). However, there are objective, mathematical criteria based on topology and thresholdsfor identifying what collections of nodes comprise amodule in alarge graphical network. Some of these metrics are described below.
