Sequence-Based Prediction of Protein-Protein Interactions Using Random Tree and Genetic Algorithm
Protein-protein interactions play important roles in the course of cell functions such as metabolic pathways and genetic information processing. There are many shortcomings of traditional experiments such as tediousness and laboriousness. The machine learning methods have been developed to predict PPIs, and preliminary results have demonstrated their feasibility. Here, we introduce a sequence-based random tree and GA to infer PPI. Experimental results on S.cerevisiae dataset from DIP show that our novel method performs well than rotation forest, with higher accuracy, sensitivity and precision. Most importantly, our method runs faster than rotation forest.
KeywordsProtein-protein interactions protein sequence autocorrelation descriptor random tree GA
Unable to display preview. Download preview PDF.
- 7.Ben-Hur, A., Noble, W.S.: Kernel Methods for Predicting Protein-protein Interactions. Bioinformatics 21, i38–i46 (2005)Google Scholar
- 16.Broto, P., Moreau, G., Vandicke, C.: Molecular Structures: Perception Autocorrelation Descriptor and Non-local Interactions. Neurocomputing 68, 66–70 (1984)Google Scholar
- 18.Houck, C.R., Joines, J.A., Kay, M.G.: A Genetic Algorithm for Function Optimization: a Matlab Implementation. Technical Report: NCSU-IE-TR-95-09, North Carolina State University, Raleigh, NC (2005)Google Scholar