Identification of Hot Regions in Protein-Protein Interactions Based on Detecting Local Community Structure
Hot regions can help proteins to exert their biological function and contribute to understand the molecular mechanism, which is the foundation of drug designs. In this paper, combining protein biological characteristics, a new method is proposed to predict protein hot regions. Firstly, we used support vector machine to predict the hot spots. Then, the local community structure detecting algorithm based on the identification of boundary nodes was proposed to predict the hot regions in protein-protein interactions. The experimental results demonstrate that the proposed method improves significantly the predictive accuracy and performance of protein hot regions.
KeywordsHot regions PPI Local community structure Classification
The authors thank the members of Machine Learning and Artificial Intelligence Laboratory, School of Computer Science and Technology, Wuhan University of Science and Technology, for their helpful discussion within seminars. This work was supported in part by National Natural Science Foundation of China (No. 61502356, 61273225, 61273303).
- 14.Nan, D.F., Zhang, X.L.: Prediction of hot regions in protein-protein interactions based on complex network and community detection. In: Bioinformatics and Biomedicine. pp. 17–23 (2013)Google Scholar