Title: Clustering PPI data based on Improved functional-flow model through Quantum-behaved PSO

Authors: Xiujuan Lei; Xu Huang; Lei Shi; Aidong Zhang

Addresses: College of Computer Science, Shaanxi Normal University, Xi'an, Shaanxi Province, 710062, China. ' College of Computer Science, Shaanxi Normal University, Xi'an, Shaanxi Province, 710062, China ' Department of Computer Science and Engineering, State University of New York at Buffalo, NY 14260-2000, USA. ' Department of Computer Science and Engineering, State University of New York at Buffalo, NY 14260-2000, USA

Abstract: Clustering Protein-Protein Interaction (PPI) data is a difficult problem due to its small world and scale-free characteristics. Existing clustering methods could not perform well. This paper proposes an improved functional-flow based approach through Quantum-behaved Particle Swarm Optimisation (QPSO) algorithm, which can find the optimum threshold automatically when calculating the lowest similarity between modules. We also take bridging nodes into account to improve the clustering result. The experiments on Munich Information Center for Protein Sequences (MIPS) PPI data sets show that the algorithm has better performance than functional flow method in terms of accuracy and number of matched clusters.

Keywords: PPI networks; protein-protein interaction; QPSO; quantum-behaved PSO; particle swarm optimisation; functional flow; clustering; protein sequences; bioinformatics.

DOI: 10.1504/IJDMB.2012.045545

International Journal of Data Mining and Bioinformatics, 2012 Vol.6 No.1, pp.42 - 60

Received: 06 Oct 2010
Accepted: 26 Jan 2011
Published online: 17 Dec 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article