A Hybrid ACO-Graph Entropy for Functional Modules Detection from Protein–Protein Interaction Network

Abstract

Recent high-throughput experiments have generated protein-protein interaction data on a genomic scale, yielding the complete protein-protein interaction network for several organisms. Various graph clustering algorithms have been applied to protein interaction networks for detecting protein functional modules. Although the previous algorithms are scalable and robust, their accuracy is still limited because of the complex connectivity found in protein interaction networks. The Ant Colony Optimization (ACO) Algorithm has been adapted for the protein functional module detection by modeling the problem as an optimization problem. The adapted ACO (ACO-PFMDA) has obtained feasible solution but not as magnificent as those reported in the literature. Some shortcomings were identified and addressed by proposing a Modified Ant Colony Optimization Algorithm (ACO-PFMDM), which introduces two new scheme for controlling the two main parameters of ACO to solve PFMDP. Experiments on one popular benchmark dataset namely Saccharomyces cerevisiae which taken from two popular databases DIP and MIPS has been performed. The experimental result have proved that ACO-PFMDM have improved the overall performance of protein functional module detection. The search process of ACO-PFMDM has converged effectively compared to some state-of-art algorithms. Moreover, the proposed dynamic update of the heuristic parameters based on entropy has generated high quality tours and it can guide ants toward the effective solutions space in the initial search stages.