Classification of Protein Modification Sites with Machine Learning

Jin Sun, Wenzheng Bao, Yi Cao, Yuehui Chen

Abstract: Lysine malonylation is a newly discovered type of protein post-translational modification, which plays an essential role in many biological activities. A good knowledge of malonylation sites can serve as guidance in solving a large number of biological problems, such as disease diagnosis and drug discovery. There have already been several experimental approaches to identify modification sites, but they are relatively expensive. In this work, we propose three novel machine learning models and utilizes several effective feature description methods. The model is trained based on the cross validation method named Split to Equal Validation (SEV). The experiments show that our model outperforms the others considerably.