Incorporating Secondary Features into the General form of Chou's PseAAC for Predicting Protein Structural Class

Bo      Liao; Qilin      Xiang; Dachao      Li

Abstract

Protein structure information is very useful for the confirmation of protein function. The protein structural class can provide information for protein 3D structure analysis, causing the conformation of the protein overall folding type plays a significant part in molecular biology. In this paper, we focus on the prediction of protein structural class which was based on new feature representation. We extract features from the Chou-Fasman parameter, amino acid compositions, amino acids hydrophobicity features, polarity information and pair-coupled amino acid composition. The prediction result by the Support vector machine (SVM) classifier shows that our method is better than some others.

Keywords: Binary sequence, feature representation, protein structure class, pseudo-amino acid composition(PseAAC), representation of protein sequence, support vector machine (SVM)