Protein Secondary Structure Prediction Using AutoEncoder Network and Bayes Classifier

2017 International Symposium on Application of Materials Science and Energy Materials

材料科学;能源学

Wang, Leilei^1 ; Cheng, Jinyong^1

College of Information, Qilu University of Technology (Shandong Academy of Sciences), Jinan, China^1

关键词: Auto encoders;    Bayes Classifier;    Cross validation;    Data set;    Prediction accuracy;    Protein secondary structure;    Protein secondary-structure prediction;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/322/6/062008/pdf
DOI  :  10.1088/1757-899X/322/6/062008

学科分类：材料科学（综合）

来源: IOP

【 摘 要 】

Protein secondary structure prediction is belong to bioinformatics,and it's important in research area. In this paper, we propose a new prediction way of protein using bayes classifier and autoEncoder network. Our experiments show some algorithms including the construction of the model, the classification of parameters and so on. The data set is a typical CB513 data set for protein. In terms of accuracy, the method is the cross validation based on the 3-fold. Then we can get the Q3accuracy. Paper results illustrate that the autoencoder network improved the prediction accuracy of protein secondary structure.