An Improved Prediction Method of Protein Disulfide Bond

Abstract

The paper presents a method to predict disulfide bond structure based on sample selection and Classifiers Fusion Technology. Firstly, the codes of the selected protein sequence are used as the input data of RBF neural network. Then the different sizes of the information windows were selected to construct the prediction models of disulfide bond. At last, the final prediction will be obtain from fusing different forecasting models. The result of above simulation experiments shows that the prediction model based on classifier fusion technology can greatly increase prediction accuracy of the structure of protein disulfide bond.