An Average-Degree Based Method for Protein Complexes Identification

In this paper, we propose an average-degree based cluster mining algorithm (ACM) for complexes detection in PPI networks. ACM method contains of three stages. Firstly, we make use of PPI network topology, i.e., average degree, to present a new quantitative function and then present a hierarchical algorithm to identify protein complexes. Finally, post-processing is applied to the predicted results to ensure the accuracy and reliability. We experimentally evaluate the performance of ACM using three different yeast PPI networks. Our results show that ACM is effective and reliable in detecting protein complexes.