Predicting of Oxidoreductase and Lyase Subclasses by Using Support Vector Machine

Abstract

Based on enzyme sequence, using composite vector with amino acid composition, low frequency of power spectral density, predicted secondary structure, value of autocorrelation function and motif frequency to express the information of sequence, an approach of support vector machine (SVM) for predicting 18 subclasses of oxidoreductases and 6 subclasses of lyases is proposed. By the Jackknife test, the overall success rates are 89. 9% and 95.1%, our predictive results are better than pervious results Keywords-enzyme, ¦Â-hairpin motif, ligand binding site, support vector machine, minimum redundancy maximum relevance.