Linear Support Vector Machines for HIV-1 Protease Site Detection

Linear Support Vector Machines for HIV-1 Protease Site Detection

Gök, Murat
Özcerit, Ahmet Turan

Several studies have been done for the HIV-1 protease specificity problem by applying machine learning computation techniques recently. In this work, a Linear Support Vector Machine (LSVM) technique has been applied to predict the cleavability of proteins by HIV-1 protease. We used Orthonormal Encoding (OE) extraction technique to map octopeptide sequence inputs. According to simulation outcomes, we have achieved better result, which has a rate of %91.8, compared to earlier studies to predict the cleavability of HIV-1 protease.

Conference or Workshop Item
PeerReviewed

2009-06

514