Interaction-Based Inductive Bias in Graph Neural Networks: Enhancing Protein-Ligand Binding Affinity Predictions From 3D Structures
文献类型：期刊论文
| 作者 | Yang, Ziduo2; Zhong, Weihe3; Lv, Qiujie2; Dong, Tiejun2; Chen, Guanxing2; Chen, Calvin Yu-Chian1,4,5,6 |
| 刊名 | IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
 |
| 出版日期 | 2024-12-01 |
| 卷号 | 46期号:12页码:8191-8208 |
| 关键词 | Proteins Programmable logic arrays Three-dimensional displays Predictive models Graph neural networks Data models Convolution Protein-ligand binding affinity graph neural networks inductive bias drug-target interaction structure-based virtual screening |
| ISSN号 | 0162-8828 |
| DOI | 10.1109/TPAMI.2024.3400515 |
| 通讯作者 | Chen, Calvin Yu-Chian(cy@pku.edu.cn) |
| 英文摘要 | Inductive bias in machine learning (ML) is the set of assumptions describing how a model makes predictions. Different ML-based methods for protein-ligand binding affinity (PLA) prediction have different inductive biases, leading to different levels of generalization capability and interpretability. Intuitively, the inductive bias of an ML-based model for PLA prediction should fit in with biological mechanisms relevant for binding to achieve good predictions with meaningful reasons. To this end, we propose an interaction-based inductive bias to restrict neural networks to functions relevant for binding with two assumptions: 1) A protein-ligand complex can be naturally expressed as a heterogeneous graph with covalent and non-covalent interactions; 2) The predicted PLA is the sum of pairwise atom-atom affinities determined by non-covalent interactions. The interaction-based inductive bias is embodied by an explainable heterogeneous interaction graph neural network (EHIGN) for explicitly modeling pairwise atom-atom interactions to predict PLA from 3D structures. Extensive experiments demonstrate that EHIGN achieves better generalization capability than other state-of-the-art ML-based baselines in PLA prediction and structure-based virtual screening. More importantly, comprehensive analyses of distance-affinity, pose-affinity, and substructure-affinity relations suggest that the interaction-based inductive bias can guide the model to learn atomic interactions that are consistent with physical reality. As a case study to demonstrate practical usefulness, our method is tested for predicting the efficacy of Nirmatrelvir against SARS-CoV-2 variants. EHIGN successfully recognizes the changes in the efficacy of Nirmatrelvir for different SARS-CoV-2 variants with meaningful reasons. |
| WOS关键词 | SCORING FUNCTIONS ; ACCURATE DOCKING ; DATABASE ; GLIDE |
| 资助项目 | National Natural Science Foundation of China[62176272] ; Research and Development Program of Guangzhou Science and Technology Bureau[2023B01J1016] ; Key-Area Research and Development Program of Guangdong Province[2020B1111100001] ; China Postdoctoral Science Foundation[2023M743926] |
| WOS研究方向 | Computer Science ; Engineering |
| 语种 | 英语 |
| WOS记录号 | WOS:001364431200047 |
| 出版者 | IEEE COMPUTER SOC |
| 源URL | [http://119.78.100.183/handle/2S10ELR8/314974]  |
| 专题 | 中国科学院上海药物研究所 |
| 通讯作者 | Chen, Calvin Yu-Chian |
| 作者单位 | 1.China Med Univ Hosp, Dept Med Res, Taichung 404, Taiwan 2.Sun Yat Sen Univ, Sch Intelligent Syst Engn, Artificial Intelligence Med Res Ctr, Shenzhen Campus, Shenzhen 518107, Peoples R China 3.Chinese Acad Sci, Shanghai Inst Mat Med, Zhongshan Inst Drug Discovery, Zhongshan 528400, Peoples R China 4.Peking Univ, Shenzhen Grad Sch, AI Sci AI4S Preferred Program, Shenzhen 518055, Peoples R China 5.Peking Univ, Shenzhen Grad Sch, Sch Elect & Comp Engn, Shenzhen 518055, Peoples R China 6.Asia Univ, Dept Bioinformat & Med Engn, Taichung 413, Taiwan |
| 推荐引用方式 GB/T 7714 | Yang, Ziduo,Zhong, Weihe,Lv, Qiujie,et al. Interaction-Based Inductive Bias in Graph Neural Networks: Enhancing Protein-Ligand Binding Affinity Predictions From 3D Structures[J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE,2024,46(12):8191-8208. |
| APA | Yang, Ziduo,Zhong, Weihe,Lv, Qiujie,Dong, Tiejun,Chen, Guanxing,&Chen, Calvin Yu-Chian.(2024).Interaction-Based Inductive Bias in Graph Neural Networks: Enhancing Protein-Ligand Binding Affinity Predictions From 3D Structures.IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE,46(12),8191-8208. |
| MLA | Yang, Ziduo,et al."Interaction-Based Inductive Bias in Graph Neural Networks: Enhancing Protein-Ligand Binding Affinity Predictions From 3D Structures".IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 46.12(2024):8191-8208. |
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。