Enhancing Generalizability in Protein-Ligand Binding Affinity Prediction with Multimodal Contrastive Learning

被引:1
|
作者
Luo, Ding [1 ,2 ]
Liu, Dandan [1 ,2 ]
Qu, Xiaoyang [3 ,4 ]
Dong, Lina [1 ,2 ]
Wang, Binju [1 ,2 ,5 ]
机构
[1] Xiamen Univ, Coll Chem & Chem Engn, State Key Lab Phys Chem Solid Surfaces, Xiamen 361005, Peoples R China
[2] Xiamen Univ, Coll Chem & Chem Engn, Fujian Prov Key Lab Theoret & Computat Chem, Xiamen 361005, Peoples R China
[3] Putian Univ, Sch Pharm & Med Technol, Putian 351100, Peoples R China
[4] Putian Univ, Fujian Prov Univ, Key Lab Pharmaceut Anal & Lab Med, Putian 351100, Peoples R China
[5] Innovat Lab Sci & Technol Energy Mat Fujian Prov I, Xiamen 361005, Peoples R China
基金
中国国家自然科学基金;
关键词
CONVOLUTIONAL NEURAL-NETWORK; EMPIRICAL SCORING FUNCTIONS; FREE-ENERGY CALCULATIONS; ACCURATE DOCKING; DRUG DISCOVERY; EFFICIENT; FORCE; GLIDE; MODEL; BENCHMARKING;
D O I
10.1021/acs.jcim.3c01961
中图分类号
R914 [药物化学];
学科分类号
100701 ;
摘要
Improving the generalization ability of scoring functions remains a major challenge in protein-ligand binding affinity prediction. Many machine learning methods are limited by their reliance on single-modal representations, hindering a comprehensive understanding of protein-ligand interactions. We introduce a graph-neural-network-based scoring function that utilizes a triplet contrastive learning loss to improve protein-ligand representations. In this model, three-dimensional complex representations and the fusion of two-dimensional ligand and coarse-grained pocket representations converge while distancing from decoy representations in latent space. After rigorous validation on multiple external data sets, our model exhibits commendable generalization capabilities compared to those of other deep learning-based scoring functions, marking it as a promising tool in the realm of drug discovery. In the future, our training framework can be extended to other biophysical- and biochemical-related problems such as protein-protein interaction and protein mutation prediction.
引用
收藏
页码:1892 / 1906
页数:15
相关论文
共 50 条
  • [1] Prediction of protein-ligand binding affinity with deep learning
    Wang, Yuxiao
    Jiao, Qihong
    Wang, Jingxuan
    Cai, Xiaojun
    Zhao, Wei
    Cui, Xuefeng
    [J]. COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL, 2023, 21 : 5796 - 5806
  • [2] Surface-based multimodal protein-ligand binding affinity prediction
    Xu, Shiyu
    Shen, Lian
    Zhang, Menglong
    Jiang, Changzhi
    Zhang, Xinyi
    Xu, Yanni
    Liu, Juan
    Liu, Xiangrong
    [J]. BIOINFORMATICS, 2024, 40 (07)
  • [3] Protein-Ligand Binding Affinity Prediction Based on Deep Learning
    Lu, Yaoyao
    Liu, Junkai
    Jiang, Tengsheng
    Guan, Shixuan
    Wu, Hongjie
    [J]. INTELLIGENT COMPUTING THEORIES AND APPLICATION, ICIC 2022, PT II, 2022, 13394 : 310 - 316
  • [4] SadNet: a novel multimodal fusion network for protein-ligand binding affinity prediction
    Hong, Qiansen
    Zhou, Guoqiang
    Qin, Yuke
    Shen, Jun
    Li, Haoran
    [J]. PHYSICAL CHEMISTRY CHEMICAL PHYSICS, 2024, 26 (16) : 12880 - 12891
  • [5] Deep Learning in Drug Design: Protein-Ligand Binding Affinity Prediction
    Rezaei, Mohammad A.
    Li, Yanjun
    Wu, Dapeng
    Li, Xiaolin
    Li, Chenglong
    [J]. IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS, 2022, 19 (01) : 407 - 417
  • [6] Prediction of protein-ligand binding affinity via deep learning models
    Wang, Huiwen
    [J]. BRIEFINGS IN BIOINFORMATICS, 2024, 25 (02)
  • [7] DeepAtom: A Framework for Protein-Ligand Binding Affinity Prediction
    Li, Yanjun
    Rezaei, Mohammad A.
    Li, Chenglong
    Li, Xiaolin
    [J]. 2019 IEEE INTERNATIONAL CONFERENCE ON BIOINFORMATICS AND BIOMEDICINE (BIBM), 2019, : 303 - 310
  • [8] DLSSAffinity: protein-ligand binding affinity prediction via a deep learning model
    Wang, Huiwen
    Liu, Haoquan
    Ning, Shangbo
    Zeng, Chengwei
    Zhao, Yunjie
    [J]. PHYSICAL CHEMISTRY CHEMICAL PHYSICS, 2022, 24 (17) : 10124 - 10133
  • [9] Improving the prediction of protein-ligand binding affinity using deep learning models
    Rezaei, Mohammad
    Li, Yanjun
    Li, Xiaolin
    Li, Chenglong
    [J]. ABSTRACTS OF PAPERS OF THE AMERICAN CHEMICAL SOCIETY, 2019, 257
  • [10] Development and evaluation of a deep learning model for protein-ligand binding affinity prediction
    Stepniewska-Dziubinska, Marta M.
    Zielenkiewicz, Piotr
    Siedlecki, Pawel
    [J]. BIOINFORMATICS, 2018, 34 (21) : 3666 - 3674