Improving discrimination of outer membrane proteins by fusing different forms of pseudo amino acid composition |
| |
Authors: | Qing-Bin Gao |
| |
Affiliation: | Department of Health Statistics, Second Military Medical University, No. 800 Xiangyin Road, Shanghai 200433, China |
| |
Abstract: | Integral membrane proteins are central to many cellular processes and constitute approximately 50% of potential targets for novel drugs. However, the number of outer membrane proteins (OMPs) present in the public structure database is very limited due to the difficulties in determining structure with experimental methods. Therefore, discriminating OMPs from non-OMPs with computational methods is of medical importance as well as genome sequencing necessity. In this study, some sequence-derived structural and physicochemical features of proteins were incorporated with amino acid composition to discriminate OMPs from non-OMPs using support vector machines. The discrimination performance of the proposed method is evaluated on a benchmark dataset of 208 OMPs, 673 globular proteins, and 206 α-helical membrane proteins. A high overall accuracy of 97.8% was observed in the 5-fold cross-validation test. In addition, the current method distinguished OMPs from globular proteins and α-helical membrane proteins with overall accuracies of 98.2 and 96.4%, respectively. The prediction performance is superior to the state-of-the-art methods in the literature. It is anticipated that the current method might be a powerful tool for the discrimination of OMPs. |
| |
Keywords: | Membrane protein Structure Physicochemical property Pseudo amino acid composition Computational biology |
本文献已被 ScienceDirect 等数据库收录! |
|