Nphos: Database and Predictor of ProteinN-phosphorylation

Abstract

AbstractProteinN-phosphorylation widely present in nature and participates in various biological functions. However, current knowledge onN-phosphorylation is extremely limited compared to that onO-phosphorylation. In this study, we collected 11,710 experimentally verifiedN-phosphosites of 7344 proteins from 39 species and subsequently constructed the database Nphos to share up-to-date information on proteinN-phosphorylation. Upon these substantial data, we characterized the sequential and structural features of proteinN-phosphorylation. Moreover, after comparing of hundreds of learning models, we chose and optimized gradient boosting decision tree (GBDT) models to predict three types of humanN-phosphorylation, achieving mean areas under the receiver operating characteristic curve (AUC) of 90.56%, 91.24%, and 92.01% for pHis, pLys, and pArg, respectively. Meanwhile, we discovered 488,825 distinctN-phosphosites in the human proteome. The models were also deployed in Nphos for interactiveN-phosphosite prediction. In summary, this work provides new insights and points for both flexible and focused investigations ofN-phosphorylation. It will also facilitate a deeper and more systematic understanding of proteinN-phosphorylation modification by providing a data and technical foundation. Nphos is freely available athttp://www.bio-add.organdhttp://ppodd.org.cn/Nphos/.