KSIMC: Predicting Kinase⁻Substrate Interactions Based on Matrix Completion.

Protein phosphorylation is an important chemical modification catalyzed by kinases. It plays important roles in many cellular processes. Predicting kinase⁻substrate interactions is vital to understanding the mechanism of many diseases. Many computational methods have been proposed to identify kinase⁻substrate interactions. However, the prediction accuracy still needs to be improved. Therefore, it is necessary to develop an efficient computational method to predict kinase⁻substrate interactions. In this paper, we propose a novel computational approach, KSIMC, to identify kinase⁻substrate interactions based on matrix completion. Firstly, the kinase similarity and substrate similarity are calculated by aligning sequence of kinase⁻kinase and substrate⁻substrate, respectively. Then, the original association network is adjusted based on the similarities. Finally, the matrix completion is used to predict potential kinase⁻substrate interactions. The experiment results show that our method outperforms other state-of-the-art algorithms in performance. Furthermore, the relevant databases and scientific literature verify the effectiveness of our algorithm for new kinase⁻substrate interaction identification.