Title: A supervised approach to detect protein complex by combining biological and topological properties

Authors: Yang Yu; Xiaolong Wang; Lei Lin; Chengjie Sun; Xuan Wang

Addresses: Department of Computer Science and Technology, Harbin Institute of Technology Shenzhen Graduate School, Shenzhen 518055, China ' School of Computer Science and Technology, Harbin Institute of Technology, Harbin, 150001 China; Department of Computer Science and Technology, Harbin Institute of Technology Shenzhen Graduate School, Shenzhen 518055, China ' School of Computer Science and Technology, Harbin Institute of Technology, Harbin, 150001 China ' School of Computer Science and Technology, Harbin Institute of Technology, Harbin 150001, China ' Department of Computer Science and Technology, Harbin Institute of Technology Shenzhen Graduate School, Shenzhen 518055, China

Abstract: Protein biochemical functions are revealed in the form of protein complexes from the Protein-Protein Interaction (PPI) network. In this paper, we propose a supervised based algorithm by combining biological and topological properties to detect protein complexes in PPI networks, in which protein amino acid background frequency is introduced as biological properties and become a new block of the features. In comparison with other established methods, the evaluation results indicate that the applied method can achieve comparable performances and match more meaningful real complexes. We also demonstrate that the use of protein amino acid background frequency and the SVM based method can efficiently improve the performance.

Keywords: PPI; protein-protein interaction; protein complex detection; amino acid background frequency; SVM; support vector machines; biological properties; topological properties; bioinformatics; protein complexes.

DOI: 10.1504/IJDMB.2013.054700

International Journal of Data Mining and Bioinformatics, 2013 Vol.8 No.1, pp.105 - 121

Received: 19 Mar 2011
Accepted: 11 Oct 2011
Published online: 20 Oct 2014 *

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