Prediction of Phosphorylation Sites Using PSO-ANNs

• Published in: Intelligent Computing Theories and Application Vol. 9771; pp. 347 - 355
• Main Authors: Han, Ruizhi, Wang, Dong, Chen, Yuehui, Bao, Wenzheng, Zhang, Qianqian, Cong, Hanhan
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2016; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Artificial intelligence; Artificial neural network; Particle swarm optimization (PSO); Phosphorylation sites prediction
• ISBN: 9783319422909; 3319422901
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-319-42291-6_34

Post-translational modifications (PTMs) are essential for regulating conformational changes, activities and functions of proteins, and are involved in almost all cellular pathways and processes. Phosphorylation is one of the most important post-translational modifications of proteins, which is related to many activities of life. It can regulate signal transduction, gene expression and cell cycle regulation of many cellular processes by protein phosphorylation and dephosphorylation. With the development and application of proteomics technology, researchers pay close attention on protein phosphorylation research more and more widely. In this paper, we use PSO algorithm to optimize neural network weight coefficients and classify the data which has secondary encoding according to the physical and chemical properties of amino acids for feature extraction. The experimental results compared with the result of the support vector machine (SVM) and experimental results show that the prediction accuracy of PSO-ANNs 2.44 % higher than that of SVM. And this paper at the same time, this paper also analyzes the experimental results under different window values. The results of the experiment are best when the window value is 11.