Phosphorylation site prediction

Phosphorylation is one of the most important posttranslational modification events for the regulation and maintenance of most biological processes. The computational prediction of phosphorylation sites is widely used as a complementary strategy for the fast and coarse determination of candidate sites. During the past 20 years, many prediction algorithms and tools have been proposed to solve this problem. Essentially, the problem of phosphorylation site prediction is cast as a classification problem in data mining and machine learning. This chapter provides a summary of the data preprocessing step, with a particular emphasis on the training data set construction. Then, we introduce some representative learning schemes that have been adopted to model the phosphorylation site prediction problem. Finally, we discuss the limitations of current solutions and highlight future research directions.