Author/Authors :
Gök, Murat Yalova University - Department of Computer Engineering, Turkey , Özcerit, Ahmet Turan Sakarya University - Department of Computer Electronics, Turkey
Title Of Article :
Proteomic Analysis of HIV-1 Protease Specificity With A New Feature Encoding Method
Abstract :
In this study a new feature encoding scheme named FTKY (Physicochemical Based Encoding Method) has been developed for HIV-1 protease site prediction. FTKY has been tested according to selected best 10-pc(physicochemical), 20-pc, 30-pc, 40-pc and 50-pc by means of accuracy, specificity and AUROC (Area Under Receiver Operating Charecteristic Curve) on Linear Support Vector Machines. Tests have been conducted on two up-to-date HIV-1 protease datasets, PR-1625 and PR-3265. According to empirical results, FTKY has been performed better prediction of accuracy 95.21 % on PR-1625 according to best 10-pc and accuracy 94.37 % when using PR-3261 according to best 30-pc on a standalone classifier.
NaturalLanguageKeyword :
HIV , 1 Protease Specificity , Feature Representation , Peptide Classification
JournalTitle :
Erciyes University Journal Of The Institute Of Science and Technology
