Identifying protein complexes based on neighborhood density in weighted PPI networks

Abstract

Most proteins form macromolecular complexes to perform their biological functions. With the increasing availability of large amounts of high-throughput protein-protein interaction (PPI) data, a vast number of computational approaches for detecting protein complexes have been proposed to discover protein complexes from PPI networks. However, such approaches are not good enough since the high rate of noise in high-throughput PPI data, including spurious and missing interactions. In this paper, we present an algorithm for complexes identification based on neighborhood density (CIND) in weighted PPI networks. Firstly, we assigned each binary protein interaction a weight, reflecting the confidence that this interaction is a true positive interaction. Then we identify complexes based on neighborhood density using topological, and we should put attention to not only the very dense regions but also the regions with low neighborhood density. We experimentally evaluate the performance of our algorithm CIND on a few yeast PPI networks, and show that our algorithm is able to identify complexes more accurately than existing algorithms.