An improved AP algorithm for identifying overlapping functional modules in protein-protein interaction networks

Identifying modules in protein-protein interaction (PPI) networks is important to understand the organization of the cellular processes. In this paper, an improved algorithm based on affinity propagation (AP) is proposed. We embed AP in our algorithm by utilizing AP to find the candidate overlapping vertices and keep those satisfying our filter condition. We apply our algorithm to S. cerevisiae PPI networks. The experimental results show that compared with AP, MCL, MCODE and CPM, our algorithm can discover more functional modules with high matching rate. Our proposed method is validated as an effective algorithm in identifying overlapping functional modules and can provide more insights for future biological study.