Medicine (SJTUSM)

^ These authors are considered to be the joint first authors * Corresponding authors Abstract As an important covalent post-translational modification, Tyrosine nitration is closely related to the causes of various diseases. However, its mechanism is still largely unknown. Based on experimentally verified tyrosine nitration sites, we introduced a novel computational approach to analyze tyrosine nitration. Nearest Neighbor algorithm armed by maximum relevancy minimum redundancy feature selection approach was used to construct the tyrosine nitration predictor. The problem of the imbalance of dataset sizes was dealt with by dividing large dataset into several smaller ones. Analysis of the selected features shows some interesting phenomena of the tyrosine nitration, which could be helpful for further computational and experimental investigations