计算机应用 ›› 2017, Vol. 37 ›› Issue (5): 1512-1515.DOI: 10.11772/j.issn.1001-9081.2017.05.1512

• 应用前沿、交叉与综合 • 上一篇    下一篇

基于深度学习的八类蛋白质二级结构预测算法

张蕾, 李征, 郑逢斌, 杨伟   

  1. 河南大学 计算机与信息工程学院, 河南 开封 475004
  • 收稿日期:2016-10-28 修回日期:2016-12-02 出版日期:2017-05-10 发布日期:2017-05-16
  • 通讯作者: 杨伟
  • 作者简介:张蕾(1983-),女,河南周口人,助教,硕士,主要研究方向:生物信息学;李征(1985-),女,河南驻马店人,讲师,博士,主要研究方向:软件工程;郑逢斌(1963-),男,河南信阳人,教授,博士,主要研究方向:空间信息处理、自然语言处理;杨伟(1983-),男,河南信阳人,讲师,博士,主要研究方向:机器学习、深度学习。
  • 基金资助:
    国家自然科学基金面上项目(41571417)。

Prediction of eight-class protein secondary structure based on deep learning

ZHANG Lei, LI Zheng, ZHENG Fengbin, YANG Wei   

  1. School of Computer and Information Engineering, Henan University, Kaifeng Henan 475004, China
  • Received:2016-10-28 Revised:2016-12-02 Online:2017-05-10 Published:2017-05-16
  • Supported by:
    This work is partially supported by the National Natural Science Foundation of China (41571417).

摘要: 蛋白质二级结构预测是结构生物学中的一个重要问题。针对八类蛋白质二级结构预测,提出了一种基于递归神经网络和前馈神经网络的深度学习预测算法。该算法通过双向递归神经网络建模氨基酸间的局部和长程相互作用,递归神经网络的隐层输出进一步送入到三层的前馈神经网络以便进行八类蛋白质二级结构预测。实验结果表明,提出的算法在CB513数据集上达到了67.9%的Q8预测精度,显著地优于SSpro8和SC-GSN。

关键词: 深度学习, 递归神经网络, 前馈神经网络, 蛋白质二级结构预测

Abstract: Predicting protein secondary structure is an important issue in structural biology. Aiming at the prediction of eight-class protein secondary structure, a novel deep learning prediction algorithm was proposed by combining recurrent neural network and feed-forward neural network. A bidirectional recurrent neural network was used to model locality and long-range interaction between amino acid residues in protein. In order to predict the eight-class protein secondary structure, the outputs of the hidden layer in the bidirectional recurrent neural network were further fed to the three-layer feed-forward neural network. Experimental results show that the proposed method achieves Q8 accuracy of 67.9% on the CB513 dataset, which is significantly better than SSpro8 and SC-GSN (Supervised Convolutional-Generative Stochastic Network).

Key words: deep learning, recurrent neural network, feed-forward neural network, protein secondary structure prediction

中图分类号: