A Multiobjective Go based Approach to Protein Complex Detection

Considering the large number of proteins and the high complexity of the protein interaction network, decomposing it into smaller and manageable modules is a valuable step toward the analysis of biological processes and pathways. There are di_erent types of graph clustering approaches available for modularization of PPI networks. But most of them seek to identify groups or clusters of proteins on the basis of the local density of the interaction graph in the network. So they find the denser region in the interaction graph, as the denser regions are expected to have the high chance to form a protein complex. In this article we present a novel Multiobjective Gene Ontology based Clustering (MGOC) in PPI network- to find protein complexes. We consider both graphical properties of PPI network as well as biological properties based on GO semantic similarity as objective functions. The proposed technique is demonstrated on yeast PPI data and the results are compared with that of some existing algorithms.