Detecting Protein Complexes Using Multiple Node Property of PPI Networks

Identifying protein complexes associated with different cellular activities have always fascinated researchers since decades. Many works have been proposed which deals with identification of protein complexes based on various criterion such as connectivity, clustering coefficient or other topological measures. Some other methods use a combination of functional similarity apart from network topology attribute to identify protein complexes. In this work, we have proposed a technique to detect protein complexes using the concept of multi node clustering coefficient and hence we have named it as MnCC method. This technique solely depends on the network feature and uses a seed expansion strategy to detect complexes. The performance of MnCC has been compared with state-of-art algorithms in terms of Sensitivity, Positive Predictive Value and Accuracy for human PPI network. It has been found that the performance of our proposed MnCC is at-par with the contemporary methods available for protein complex identification in case of human PPI dataset.