Community structure in "gene-to-gene" networks

Girvan-Newman (GN): This popular method utilizes the idea that edges between communities should have many shortest paths crossing them. It removes those edges succedingly, which causes the network to split into parts. The splitting obtained in this way corresponding to the highest modularity is then the communities. The algorithm is generally believed to yield decent Q, however its time complexity O(# links) makes it infeasible for networks of thousands of nodes.