Detection of Locally Over-Represented GO Terms in Protein-Protein Interaction Networks

摘要: High-throughput methods for identifying protein-protein interactions produce increasingly complex and intricate interaction networks. These networks are extremely rich in information, but extracting biologically meaningful hypotheses from them representing a human-readable manner is challenging. We propose method to identify Gene Ontology terms that locally over-represented subnetwork of given biological network. Specifically, we two evaluate the degree clustering proteins associated particular GO term describe four efficient estimate statistical significance observed clustering. show, using Monte Carlo simulations, our best approximation accurately true p-value, random scale-free graphs as well actual yeast human When applied these networks, approach recovers many known complexes pathways, also suggests potential functions subnetworks.