Title: Protein contact map prediction using committee machine approach

Authors: Narjeskhatoon Habibi; Mohamad Saraee; Hassan Korbekandi

Addresses: Department of Electrical and Computer Engineering, Isfahan University of Technology, Isfahan 84156-83111, Iran ' School of Computing, Science and Engineering, University of Salford, Manchester M5 4WT, UK ' Department of Genetics and Molecular Biology, Faculty of Medicine, Isfahan University of Medical Sciences, Isfahan 81744-176, Iran

Abstract: Protein contact map is a simplified representation of a protein's spatial structure. The Committee Machine is a machine learning method that allots the learning task to a number of learners and divides the input space into subspaces. Learners' responses to an input are combined to produce the system's final response, which is more accurate than any single individual's response. In this study, we propose a novel method called CMP_model, for contact map prediction based on the committee machine. The results of the proposed model in comparison with two other models, show considerable gain (an accuracy improvement from 0.05 to 0.15).

Keywords: bioinformatics; artificial intelligence; machine learning; committee machine; ensemble learning; data mining; neural networks; protein contact maps; contact map prediction.

DOI: 10.1504/IJDMB.2013.054226

International Journal of Data Mining and Bioinformatics, 2013 Vol.7 No.4, pp.397 - 415

Received: 29 Sep 2010
Accepted: 03 Oct 2011
Published online: 20 Oct 2014 *

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