CeFunMO: A centrality based method for discovering functional motifs with application in biological networks

Abstract

Detecting functional motifs in biological networks is one of the challenging problems in systems biology. Given a multiset of colors as query and a list-colored graph (an undirected graph with a set of colors assigned to each of its vertices), the problem is reduced to finding connected subgraphs, which best cover the multiset of query. To solve this NP-complete problem, we propose a new color-based centrality measure for list-colored graphs. Based on this newly-defined measure of centrality, a novel polynomial time algorithm is developed to discover functional motifs in list-colored graphs, using a greedy strategy. This algorithm, called CeFunMO, has superior running time and acceptable accuracy in comparison with other well-known algorithms, such as RANGI and GraMoFoNe.

Keywords: Biological network; Centrality; Functional motif; List-colored graph; Protein complex.