CNN-based two-branch multi-scale feature extraction network for retrosynthesis prediction.

Journal: BMC bioinformatics
Published Date:

Abstract

BACKGROUND: Retrosynthesis prediction is the task of deducing reactants from reaction products, which is of great importance for designing the synthesis routes of the target products. The product molecules are generally represented with some descriptors such as simplified molecular input line entry specification (SMILES) or molecular fingerprints in order to build the prediction models. However, most of the existing models utilize only one molecular descriptor and simply consider the molecular descriptors in a whole rather than further mining multi-scale features, which cannot fully and finely utilizes molecules and molecular descriptors features.

Authors

  • Feng Yang
  • Juan Liu
    Key State Laboratory of Software Engineering, School of Computer, Wuhan University, Wuhan 430072, PR China. Electronic address: liujuan@whu.edu.cn.
  • Qiang Zhang
    Yunan Provincial Center for Disease Control and Prevention, Kunming 650022, China.
  • Zhihui Yang
    Institute of Artificial Intelligence, School of Computer Science, Wuhan University, China. Electronic address: zhy@whu.edu.cn.
  • Xiaolei Zhang
    College of Biosystems Engineering and Food Science, Zhejiang University, Hangzhou, Zhejiang, 310058, China; Key Laboratory of on Site Processing Equipment for Agricultural Products, Ministry of Agriculture and Rural Affairs, China.