Mining hub-based protein complexes in massive biological networks

Advanced technologies are producing large-scale protein-protein interaction data at an ever increasing pace. Finding protein-protein interaction complexes from large PPI networks is a fundamental problem in bioinformatics. As a group of core proteins which interacts with other more proteins, hub proteins play a key role in protein complex and life activity. In this paper, we propose a novel topological model, HP*-complex, which defines the hub proteins of protein complex and extends to encompass the neighborhood of the hub proteins, for the initial structure of protein complexes. An algorithm based on the new topological model, called HPCMiner, is developed for identifying protein complexes from large PPI networks. The experiment results on real dataset show that our proposed algorithm detects many complexes having special biological significance. The results from a study on synthetic data sets demonstrate that the HPCMiner algorithm scales well with respect to data set size.