Title: Weighted feature value based Drug Target Protein prediction

Authors: Bo-ra Hyun, Hwiesung Jung, Woo-Hyuk Jang, Suk Hoon Jung, Dong-Soo Han

Addresses: School of Engineering, Information and Communication University, 119, Munjiro, Yuseong-gu, Daejeon, 305-714, Korea. ' School of Engineering, Information and Communication University, 119, Munjiro, Yuseong-gu, Daejeon, 305-714, Korea. ' School of Engineering, Information and Communication University, 119, Munjiro, Yuseong-gu, Daejeon, 305-714, Korea. ' School of Engineering, Information and Communication University, 119, Munjiro, Yuseong-gu, Daejeon, 305-714, Korea. ' School of Engineering, Information and Communication University, 119, Munjiro, Yuseong-gu, Daejeon, 305-714, Korea

Abstract: Drug discovery is a long process in which only a few successful new therapeutic discoveries are made and identification of drug target candidate proteins requires considerable time and efforts. However, the accumulation of information on drugs has made it possible to devise new computational methods for classifying drug target candidates. In this paper, we devise a Drug Target Protein (DT-P) classification method by the summation of weighted features which is extracted from known DT-P. The method is validated using Bayesian decision theory and SVM, and it was revealed to achieve high specificity of 89.5% with 88% accuracy.

Keywords: drug design; DT-P; drug target proteins; weighted feature; SVM; Bayesian decision theory; drug target candidates; support vector machines; bioinformatics; drug discovery.

DOI: 10.1504/IJCBDD.2008.022211

International Journal of Computational Biology and Drug Design, 2008 Vol.1 No.4, pp.422 - 433

Published online: 22 Dec 2008 *

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