Structure-Based Prediction of Protein Phosphorylation Sites Using an Ensemble Approach

Yong Gao,Weilin Hao,Zhigang Chen,Lei Deng
DOI: https://doi.org/10.1007/978-3-319-09330-7_15
2014-01-01
Abstract:As one of the most prevailing post-translational modifications, phosphorylation is vital in regulating almost every cellular behavior. In this paper, we propose a new computational method that can effectively identify phosphorylation sites by using optimally chosen properties. The highlight of our method is that the optimal combination of features was selected from a set of 165 novel structural neighborhood properties by a random forest feature selection method. And then an ensemble learning method based on support vector machine was used to build the prediction model. Experimental results obtained from cross validation and independent test suggested that our method achieved a significant improvement on the prediction quality. Promising results were obtained after being compared with the state-of-the-art approaches using independent dataset.
What problem does this paper attempt to address?