A Combination of Feature Extraction Methods with an Ensemble of Support Vector Machines for Bacterial Virulent Proteins Prediction

作者：Liu, Baili; Zhu, Wen^*; Li, Bo; Cao, Zhi

来源：Journal of Computational and Theoretical Nanoscience, 2015, 12(8): 1813-1817.

DOI：10.1166/jctn.2015.3963

• Predicted Secondary Structure-Based Features
• Support Vector Machine
• Bacterial Virulent Protein Prediction

The availability of a reliable prediction method for prediction of bacterial virulent proteins is helpful for finding novel drug targets, vaccine candidates, and understanding virulence mechanisms in pathogens. In this study, we proposed a combination of feature extraction model which incorporates sequence-based features of amino acid sequence (dipeptide composition and Chou's Pseudo Amino Acid composition), evolutionary information-based feature(position-specific scoring matrix) and predicted secondary structure-based features simultaneously. It is the first time to use predicted secondary structure-based features to predict bacterial virulent proteins. Then we applied a bi-layer cascade SVM model to make prediction on the three most frequently used datasets and made comparison with other methods. The results on these datasets show improvements on bacterial virulent proteins prediction.

• 出版日期2015-8
• 单位湖南大学