中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 上海药物研究所
  3. 中国科学院上海药物研究所
  4. 新药研究国家重点实验室
Generic protein-ligand interaction scoring by integrating physical prior knowledge and data augmentation modelling

文献类型：期刊论文

作者Cao, Duanhua7,8; Chen, Geng5,6,7; Jiang, Jiaxin7; Yu, Jie6,7; Zhang, Runze6,7; Chen, Mingan3,4,7; Zhang, Wei6,7; Chen, Lifan6,7; Zhong, Feisheng6,7; Zhang, Yingying2,7
刊名NATURE MACHINE INTELLIGENCE
出版日期2024-06-06
页码21
DOI10.1038/s42256-024-00849-z
通讯作者Zheng, Mingyue(myzheng@simm.ac.cn)
英文摘要Developing robust methods for evaluating protein-ligand interactions has been a long-standing problem. Data-driven methods may memorize ligand and protein training data rather than learning protein-ligand interactions. Here we show a scoring approach called EquiScore, which utilizes a heterogeneous graph neural network to integrate physical prior knowledge and characterize protein-ligand interactions in equivariant geometric space. EquiScore is trained based on a new dataset constructed with multiple data augmentation strategies and a stringent redundancy-removal scheme. On two large external test sets, EquiScore consistently achieved top-ranking performance compared to 21 other methods. When EquiScore is used alongside different docking methods, it can effectively enhance the screening ability of these docking methods. EquiScore also showed good performance on the activity-ranking task of a series of structural analogues, indicating its potential to guide lead compound optimization. Finally, we investigated different levels of interpretability of EquiScore, which may provide more insights into structure-based drug design. Machine learning can improve scoring methods to evaluate protein-ligand interactions, but achieving good generalization is an outstanding challenge. Cao et al. introduce EquiScore, which is based on a graph neural network that integrates physical knowledge and is shown to have robust capabilities when applied to unseen protein targets.
WOS关键词POSE PREDICTION ; ACCURATE ; DOCKING ; MOLECULES
资助项目National Natural Science Foundation of China[T2225002] ; National Natural Science Foundation of China[82273855] ; National Natural Science Foundation of China[82204278] ; National Key Research and Development Program of China[2023YFC2305904] ; Shanghai Municipal Science and Technology Major Project ; SIMM-SHUTCM Traditional Chinese Medicine Innovation Joint Research Program[E2G805H] ; Youth Innovation Promotion Association CAS[2023296]
WOS研究方向Computer Science
语种英语
WOS记录号WOS:001242205000001
出版者NATURE PORTFOLIO
源URL[http://119.78.100.183/handle/2S10ELR8/311621]  
专题新药研究国家重点实验室
通讯作者Zheng, Mingyue
作者单位1.Nanjing Univ Chinese Med, Sch Chinese Mat Med, Nanjing, Jiangsu, Peoples R China
2.Univ Sci & Technol China, Div Life Sci & Med, Hefei, Peoples R China
3.Lingang Lab, Shanghai, Peoples R China
4.Shanghai Tech Univ, Sch Phys Sci & Technol, Shanghai, Peoples R China
5.UCAS, Hangzhou Inst Adv Study, Sch Pharmaceut Sci & Technol, Hangzhou, Peoples R China
6.Univ Chinese Acad Sci, Beijing, Peoples R China
7.Chinese Acad Sci, Shanghai Inst Mat Med, Drug Discovery & Design Ctr, State Key Lab Drug Res, Shanghai, Peoples R China
8.Zhejiang Univ, Innovat Inst Artificial Intelligence Med, Coll Pharmaceut Sci, Hangzhou, Peoples R China
推荐引用方式
GB/T 7714		 Cao, Duanhua,Chen, Geng,Jiang, Jiaxin,et al. Generic protein-ligand interaction scoring by integrating physical prior knowledge and data augmentation modelling[J]. NATURE MACHINE INTELLIGENCE,2024:21.
APA Cao, Duanhua.,Chen, Geng.,Jiang, Jiaxin.,Yu, Jie.,Zhang, Runze.,...&Zheng, Mingyue.(2024).Generic protein-ligand interaction scoring by integrating physical prior knowledge and data augmentation modelling.NATURE MACHINE INTELLIGENCE,21.
MLA Cao, Duanhua,et al."Generic protein-ligand interaction scoring by integrating physical prior knowledge and data augmentation modelling".NATURE MACHINE INTELLIGENCE (2024):21.

入库方式： OAI收割

来源：上海药物研究所

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。