Multiobjective Approach to Protein Complex Detection

Detecting protein complexes within protein-protein interaction (PPI) networks is crucial for understanding biological processes, and it remains a challenging task. This chapter focuses on identifying human protein complexes through a multiobjective framework. The large human PPI network is partitioned into modules serving as protein complexes. The framework integrates topological properties of the PPI network and biological properties based on Gene Ontology semantic similarity to build objective functions. The multiobjective method is compared with state-of-the-art algorithms using various performance metrics, and biological validation is performed through Gene Ontology and pathway analysis. An analysis associating resulting protein complexes with 22 key disease classes reveals associations with disorders like “Cancer,” “Endocrine, ” and “Multiple. ” This work presents the identification of protein complexes as a multiobjective optimization problem, uncovering potential relationships between disorders and predicted complexes that could contribute to the prediction of multi-target drugs.