AbstractAccurately predicting protein–ligand interactions is crucial for understanding cellular processes. We introduce SurfDock, a deep-learning method that addresses this challenge by integrating protein sequence, three-dimensional structural graphs and surface-level features into an equivariant architecture.

摘要准确预测蛋白质与配体的相互作用对于理解细胞过程至关重要。我们介绍了SurfDock，这是一种深度学习方法，通过将蛋白质序列，三维结构图和表面级特征整合到等变体系结构中来解决这一挑战。

SurfDock employs a generative diffusion model on a non-Euclidean manifold, optimizing molecular translations, rotations and torsions to generate reliable binding poses. Our extensive evaluations across various benchmarks demonstrate SurfDock’s superiority over existing methods in docking success rates and adherence to physical constraints.

SurfDock在非欧几里德流形上采用生成扩散模型，优化分子平移，旋转和扭转以产生可靠的结合姿势。我们对各种基准的广泛评估表明，SurfDock在对接成功率和遵守物理限制方面优于现有方法。

It also exhibits remarkable generalizability to unseen proteins and predicted apo structures, while achieving state-of-the-art performance in virtual screening tasks. In a real-world application, SurfDock identified seven novel hit molecules in a virtual screening project targeting aldehyde dehydrogenase 1B1, a key enzyme in cellular metabolism.

它还对看不见的蛋白质和预测的载脂蛋白结构表现出显着的普遍性，同时在虚拟筛选任务中实现了最先进的性能。在现实世界的应用中，SurfDock在一个针对细胞代谢关键酶醛脱氢酶1B1的虚拟筛选项目中确定了七个新的命中分子。

This showcases SurfDock’s ability to elucidate molecular mechanisms underlying cellular processes. These results highlight SurfDock’s potential as a transformative tool in structural biology, offering enhanced accuracy, physical plausibility and practical applicability in understanding protein–ligand interactions..

这展示了SurfDock阐明细胞过程潜在分子机制的能力。这些结果突出了SurfDock作为结构生物学变革工具的潜力，在理解蛋白质-配体相互作用方面提供了更高的准确性，物理合理性和实用性。。

Access through your institution

通过您的机构访问

Buy or subscribe

购买或订阅

This is a preview of subscription content, access via your institution

这是订阅内容的预览，可通过您的机构访问

Access options

访问选项

Access through your institution

通过您的机构访问

Access through your institution

通过您的机构访问

Change institution

变革机构

Buy or subscribe

购买或订阅

Access Nature and 54 other Nature Portfolio journals

Access Nature和54种其他Nature投资组合期刊

Get Nature+, our best-value online-access subscription

获取Nature+，我们最具价值的在线访问订阅

24,99 € / 30 days

24,99欧元/30天

cancel any time

随时取消

Learn more

了解更多信息

Subscription info for Chinese customersWe have a dedicated website for our Chinese customers. Please go to naturechina.com to subscribe to this journal.Go to naturechina.com

中国客户的订阅信息我们为中国客户提供了一个专门的网站。请访问naturechina.com订阅本期刊。访问naturechina.com

Buy this article

购买这篇文章

Purchase on SpringerLink

在SpringerLink上购买

Instant access to full article PDF

即时访问全文PDF

Buy now

立即购买

Prices may be subject to local taxes which are calculated during checkout

价格可能需要缴纳结帐时计算的地方税

Additional access options:

其他访问选项：

Log in

登录

Learn about institutional subscriptions

了解机构订阅

Read our FAQs

阅读我们的常见问题

Contact customer support

联系客户支持

Fig. 1: The overall architecture of SurfDock.Fig. 2: Comparative performance of docking methods across benchmarks.Fig. 3: Evaluation of SurfDock’s sampling efficiency and SurfScore’s ranking ability.Fig. 4: The performance of different docking methods on the DEKOIS 2.0 dataset.Fig. 5: SurfDock identified new scaffold hits for ALDH1B1..

图1:SurfDock的整体架构。图2：跨基准的对接方法的比较性能。图3:SurfDock采样效率和SurfScore排名能力的评估。图4:DEKOIS 2.0数据集上不同对接方法的性能。图5:SurfDock确定了ALDH1B1的新支架命中。。

Data availability

数据可用性

The data included in our paper are all from public datasets. Source data are provided with this paper.

本文中包含的数据均来自公共数据集。本文提供了源数据。

Code availability

代码可用性

The code used to generate the results shown in this study is available under an MIT License via GitHub at https://github.com/CAODH/SurfDock and via Zenodo at https://zenodo.org/records/13933663 (ref. 90).

用于生成本研究所示结果的代码可通过GitHub在麻省理工学院许可下获得https://github.com/CAODH/SurfDock通过Zenodohttps://zenodo.org/records/13933663（参考文献90）。

ReferencesAbramson, J. et al. Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature 630, 493–500 (2024).Article

参考文献Abramson，J。等人。与AlphaFold 3的生物分子相互作用的精确结构预测。。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Goodey, N. M. & Benkovic, S. J. Allosteric regulation and catalysis emerge via a common route. Nat. Chem. Biol. 4, 474–482 (2008).Article

Goodey，N.M。＆Benkovic，S.J。变构调节和催化通过共同的途径出现。自然化学。生物学4474-482（2008）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Noskov, S. Y., Bernèche, S. & Roux, B. Control of ion selectivity in potassium channels by electrostatic and dynamic properties of carbonyl ligands. Nature 431, 830–834 (2004).Article

Noskov，S.Y.，Bernèche，S。＆Roux，B。通过羰基配体的静电和动力学性质控制钾通道中的离子选择性。自然431830-834（2004）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Kuntz, I. D., Blaney, J. M., Oatley, S. J., Langridge, R. & Ferrin, T. E. A geometric approach to macromolecule-ligand interactions. J. Mol. Biol. 161, 269–288 (1982).Article

Kuntz，I.D.，Blaney，J.M.，Oattley，S.J.，Langridge，R。＆Ferrin，T.E。大分子-配体相互作用的几何方法。J、 分子生物学。161269-288（1982）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Blundell, T. L. Structure-based drug design. Nature 384, 23 (1996).CAS

Blundell，T.L。基于结构的药物设计。自然384,23（1996）。中科院

PubMed

PubMed

Google Scholar

谷歌学者

Kitchen, D. B., Decornez, H., Furr, J. R. & Bajorath, J. Docking and scoring in virtual screening for drug discovery: methods and applications. Nat. Rev. Drug Discov. 3, 935–949 (2004).Article

Kitchen，D.B.，Decornez，H.，Furr，J.R。＆Bajorath，J。药物发现虚拟筛选中的对接和评分：方法和应用。《药物目录》修订版。3935-949（2004）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Isert, C., Atz, K. & Schneider, G. Structure-based drug design with geometric deep learning. Curr. Opin. Struct. Biol. 79, 102548 (2023).Article

Isert，C.，Atz，K。＆Schneider，G。基于结构的药物设计与几何深度学习。货币。奥平。结构。生物学79102548（2023）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Méndez-Lucio, O., Ahmad, M., del Rio-Chanona, E. A. & Wegner, J. K. A geometric deep learning approach to predict binding conformations of bioactive molecules. Nat. Mach. Intell. 3, 1033–1039 (2021).Article

Méndez-Lucio，O.，Ahmad，M.，del Rio Chanona，E.A。＆Wegner，J.K。一种预测生物活性分子结合构象的几何深度学习方法。自然马赫数。因特尔。31033-1039（2021）。文章

Google Scholar

谷歌学者

Shen, C. et al. Boosting protein–ligand binding pose prediction and virtual screening based on residue-atom distance likelihood potential and graph transformer. J. Med. Chem. 65, 10691–10706 (2022).Article

Shen，C.等人。基于残基原子距离似然势和图形变换的蛋白质-配体结合姿势预测和虚拟筛选。J、 医学化学。6510691-10706（2022）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Ragoza, M., Hochuli, J., Idrobo, E., Sunseri, J. & Koes, D. R. Protein–ligand scoring with convolutional neural networks. J. Chem. Inf. Model. 57, 942–957 (2017).Article

。J、 化学。Inf.模型。57942-957（2017）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Tran-Nguyen, V.-K., Junaid, M., Simeon, S. & Ballester, P. J. A practical guide to machine-learning scoring for structure-based virtual screening. Nat. Protoc. 18, 3460–3511 (2023).Article

Tran Nguyen，V.-K.，Junaid，M.，Simeon，S。和Ballester，P.J。基于结构的虚拟筛选的机器学习评分实用指南。自然协议。。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Topol, E. J. High-performance medicine: the convergence of human and artificial intelligence. Nat. Med. 25, 44–56 (2019).Article

Topol，E.J。高性能医学：人类和人工智能的融合。《自然医学》25,44-56（2019）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Krishna, R. et al. Generalized biomolecular modeling and design with RoseTTAFold All-Atom. Science 384, eadl2528 (2024).Article

Krishna，R.等人。RoseTTAFold全原子的广义生物分子建模和设计。科学384，eadl2528（2024）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).Article

Jumper，J.等人。使用AlphaFold进行高度准确的蛋白质结构预测。自然596583-589（2021）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Wayment-Steele, H. K. et al. Predicting multiple conformations via sequence clustering and AlphaFold2. Nature 625, 832–839 (2024).Article

Wayment Steele，H.K.等人。通过序列聚类和AlphaFold2预测多种构象。《自然》625832-839（2024）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Baek, M. et al. Accurate prediction of protein structures and interactions using a three-track neural network. Science 373, 871–876 (2021).Article

Baek，M.等人，《使用三轨神经网络精确预测蛋白质结构和相互作用》，《科学》373871–876（2021）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Lin, Z. et al. Evolutionary-scale prediction of atomic-level protein structure with a language model. Science 379, 1123–1130 (2023).Article

Lin，Z.等人。用语言模型预测原子级蛋白质结构的进化规模。科学3791123-1130（2023）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Eberhardt, J., Santos-Martins, D., Tillack, A. F. & Forli, S. AutoDock Vina 1.2.0: new docking methods, expanded force field, and Python bindings. J. Chem. Inf. Model. 61, 3891–3898 (2021).Article

Eberhardt，J.，Santos Martins，D.，Tillack，A.F。＆Forli，S。AutoDock Vina 1.2.0：新的对接方法，扩展的力场和Python绑定。J、 化学。Inf.模型。613891-3898（2021）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Friesner, R. A. et al. Glide: a new approach for rapid, accurate docking and scoring. 1. Method and assessment of docking accuracy. J. Med. Chem. 47, 1739–1749 (2004).Article

Friesner，R.A。等人，《Glide：一种快速、准确对接和评分的新方法》。1。对接精度的方法和评估。J、 医学化学。471739-1749（2004）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Verdonk, M. L., Cole, J. C., Hartshorn, M. J., Murray, C. W. & Taylor, R. D. Improved protein–ligand docking using GOLD. Proteins 52, 609–623 (2003).Article

Verdonk，M.L.，Cole，J.C.，Hartshorn，M.J.，Murray，C.W。和Taylor，R.D。改进了使用GOLD的蛋白质-配体对接。蛋白质52609-623（2003）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Masters, M. R., Mahmoud, A. H., Wei, Y. & Lill, M. A. Deep learning model for efficient protein–ligand docking with implicit side-chain flexibility. J. Chem. Inf. Model. 63, 1695–1707 (2023).Article

Masters，M.R.，Mahmoud，A.H.，Wei，Y。＆Lill，M.A。深度学习模型，用于有效的蛋白质-配体对接和隐式侧链柔性。J、 化学。Inf.模型。631695-1707（2023）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Koes, D. R., Baumgartner, M. P. & Camacho, C. J. Lessons learned in empirical scoring with smina from the CSAR 2011 benchmarking exercise. J. Chem. Inf. Model. 53, 1893–1904 (2013).Article

Koes，D.R.，Baumgartner，M.P。和Camacho，C.J。在2011年CSAR基准测试练习中，用smina进行经验评分的经验教训。J、 化学。Inf.模型。531893-1904（2013）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Cao, D. et al. Generic protein–ligand interaction scoring by integrating physical prior knowledge and data augmentation modelling. Nat. Mach. Intell. 6, 688–700 (2024).Article

Cao，D.等人。通过整合物理先验知识和数据增强建模进行通用蛋白质-配体相互作用评分。自然马赫数。因特尔。6688-700（2024）。文章

Google Scholar

谷歌学者

Li, H., Sze, K.-H., Lu, G. & Ballester, P. J. Machine‐learning scoring functions for structure‐based virtual screening. Wiley Interdiscip. Rev. Comput. Mol. Sci. https://doi.org/10.1002/wcms.1478 (2021).Ricci-Lopez, J., Aguila, S. A., Gilson, M. K. & Brizuela, C. A. Improving structure-based virtual screening with ensemble docking and machine learning.

Li，H.，Sze，K.-H.，Lu，G。＆Ballester，P.J。基于结构的虚拟筛选的机器学习评分函数。Wiley Interdiscip公司。版本计算机。分子科学。https://doi.org/10.1002/wcms.1478（2021年）。Ricci-Lopez，J.，Aguila，S.A.，Gilson，M.K。和Brizuela，C.A。通过集成对接和机器学习改进基于结构的虚拟筛选。

J. Chem. Inf. Model. 61, 5362–5376 (2021).Article .

J、 化学。Inf.模型。615362-5376（2021）。文章。

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Berman, H. M. et al. The Protein Data Bank. Nucleic Acids Res. 28, 235–242 (2000).Article

Berman，H.M.等人，《蛋白质数据库》。核酸研究28235-242（2000）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Xiong, G. et al. Featurization strategies for protein–ligand interactions and their applications in scoring function development. Wiley Interdiscip. Rev. Comput. Mol. Sci. https://doi.org/10.1002/wcms.1567 (2022).Watson, J. L. et al. De novo design of protein structure and function with RFdiffusion.

Xiong，G.等人。蛋白质-配体相互作用的特征化策略及其在评分功能开发中的应用。Wiley Interdiscip公司。版本计算机。分子科学。https://doi.org/10.1002/wcms.1567（2022年）。Watson，J.L.等人。用RFdiffusion从头设计蛋白质结构和功能。

Nature 620, 1089–1100 (2023).Article .

自然6201089-1100（2023）。文章。

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Sadybekov, A. V. & Katritch, V. Computational approaches streamlining drug discovery. Nature 616, 673–685 (2023).Article

Sadybekov，A.V。＆Katritch，V。计算方法简化药物发现。自然616673-685（2023）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Zhou, G. et al. Uni-Mol: a universal 3D molecular representation learning framework. In International Conference on Learning Representations (2023).Stärk, H., Ganea, O., Pattanaik, L., Barzilay, R. & Jaakkola, T. Equibind: geometric deep learning for drug binding structure prediction.

Zhou，G。et al。Uni-Mol：一个通用的3D分子表征学习框架。国际学习表征会议（2023）。Stärk，H.，Ganea，O.，Pattanaik，L.，Barzilay，R。＆Jaakkola，T。Equibind：药物结合结构预测的几何深度学习。

In International Conference on Machine Learning 162, 20503–20521 (PMLR, 2022).Zhang, Y., Cai, H., Shi, C. & Tang, J. E3Bind: an end-to-end equivariant network for protein–ligand docking. In International Conference on Learning Representations (2023).Lu, W. et al. TANKBind: trigonometry-aware neural networks for drug-protein binding structure prediction.

在国际机器学习会议16220503-20521（PMLR，2022）中。Zhang，Y.，Cai，H.，Shi，C。＆Tang，J。E3Bind：用于蛋白质-配体对接的端到端等变网络。在国际学习表征会议（2023年）上。Lu，W。等人。TANKBind：用于药物-蛋白质结合结构预测的三角感知神经网络。

In Advances in Neural Information Processing Systems 35, 7236–7249 (Curran Associates, 2022).Zhang, X. et al. Efficient and accurate large library ligand docking with KarmaDock. Nat. Comput. Sci. 3, 789–804 (2023).Article .

《神经信息处理系统进展》357236-7249（Curran Associates，2022）。Zhang，X。等人。与KarmaDock进行高效准确的大型文库配体对接。自然计算。科学。3789-804（2023）。文章。

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Corso, G., Stärk, H., Jing, B., Barzilay, R. & Jaakkola, T. S. DiffDock: diffusion steps, twists, and turns for molecular docking. In 11th International Conference on Learning Representations (2023).Ho, J., Jain, A. & Abbeel, P. Denoising diffusion probabilistic models. In Advances in Neural Information Processing Systems 33, 6840–6851 (Curran Associates, 2020).Guan, J.

Corso，G.，Stärk，H.，Jing，B.，Barzilay，R。＆Jaakkola，T.S.DiffDock：分子对接的扩散步骤，曲折和曲折。在第十一届学习表征国际会议（2023年）上。Ho，J.，Jain，A。＆Abbeel，P。去噪扩散概率模型。《神经信息处理系统进展》336840-6851（Curran Associates，2020）。关，J。

et al. 3D equivariant diffusion for target-aware molecule generation and affinity prediction. In 11th International Conference on Learning Representations (2023).Xu, M. et al. GeoDiff: a geometric diffusion model for molecular conformation generation. In International Conference on Learning Representations (2022).Hoogeboom, E., Satorras, V.

3D等变扩散用于目标感知分子的产生和亲和力预测。在第十一届学习表征国际会议（2023年）上。Xu，M.等。GeoDiff：分子构象产生的几何扩散模型。在国际学习表征会议（2022年）上。霍格·布姆（Hoogeboom），E。萨托拉斯（Satorras），V。

G., Vignac, C. & Welling, M. Equivariant diffusion for molecule generation in 3D. In International Conference on Machine Learning 162, 8867–8887 (PMLR, 2022).Corso, G. et al. Deep confident steps to new pockets: strategies for docking generalization. In International Conference on Learning Representations (2024).Kaplan, A.

G、 ，Vignac，C。＆Welling，M。用于3D中分子生成的等变扩散。在国际机器学习会议1628867-8887（PMLR，2022）中。Corso，G.等人，《通往新口袋的自信步骤：对接泛化策略》。。卡普兰，A。

L. et al. Bespoke library docking for 5-HT2A receptor agonists with antidepressant activity. Nature 610, 582–591 (2022).Article .

五十、 等。具有抗抑郁活性的5-HT2A受体激动剂的定制文库对接。自然610582-591（2022）。文章。

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Lyu, J. et al. Ultra-large library docking for discovering new chemotypes. Nature 566, 224–229 (2019).Article

Lyu，J.等人。用于发现新化学型的超大文库对接。自然566224-229（2019）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Ackloo, S. et al. CACHE (Critical Assessment of Computational Hit-finding Experiments): a public-private partnership benchmarking initiative to enable the development of computational methods for hit-finding. Nat. Rev. Chem. 6, 287–295 (2022).Article

CACHE（计算命中发现实验的关键评估）：一项公私合作的基准测试计划，旨在开发命中发现的计算方法。自然化学修订版。6287-295（2022）。文章

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Zhang, X. et al. Advancing ligand docking through deep learning: challenges and prospects in virtual screening. Acc. Chem. Res. 57, 1500–1509 (2024).Article

Zhang，X。等人。通过深度学习推进配体对接：虚拟筛选的挑战和前景。根据化学。第571500-1509号决议（2024年）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Bender, B. J. et al. A practical guide to large-scale docking. Nat. Protoc. 16, 4799–4832 (2021).Article

Bender，B.J.等人，《大规模对接实用指南》。自然协议。164799-4832（2021）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Yu, Y., Lu, S., Gao, Z., Zheng, H. & Ke, G. Do deep learning models really outperform traditional approaches in molecular docking? In International Conference on Learning Representations (2023).Alhossary, A., Handoko, S. D., Mu, Y. & Kwoh, C.-K. Fast, accurate, and reliable molecular docking with QuickVina 2.

Yu，Y.，Lu，S.，Gao，Z.，Zheng，H.＆Ke，G.深度学习模型真的优于传统的分子对接方法吗？在国际学习表征会议（2023年）上。Alhossary，A.，Handoko，S.D.，Mu，Y.＆Kwoh，C.-K。与QuickVina 2快速，准确，可靠的分子对接。

Bioinformatics 31, 2214–2216 (2015).Article .

生物信息学312214-2216（2015）。文章。

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Buttenschoen, M., Morris, G. M. & Deane, C. M. PoseBusters: AI-based docking methods fail to generate physically valid poses or generalise to novel sequences. Chem. Sci. 15, 3130 (2024).Article

Buttenschoen，M.，Morris，G.M。＆Deane，C.M。PoseBusters：基于AI的对接方法无法产生物理上有效的姿势或推广到新序列。化学。科学。153130（2024）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Harris, C. et al. PoseCheck: generative models for 3D structure-based drug design produce unrealistic poses. In NeurIPS 2023 Generative AI and Biology (GenBio) Workshop https://neurips.cc/virtual/2023/74716 (2023).Connolly, M. L. Solvent-accessible surfaces of proteins and nucleic acids.

Harris，C。et al。PoseCheck：基于3D结构的药物设计的生成模型产生不现实的姿势。在NeurIPS 2023生成人工智能和生物学（GenBio）研讨会上https://neurips.cc/virtual/2023/74716（2023年）。。

Science 221, 709–713 (1983).Article .

科学221709-713（1983）。文章。

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Atz, K., Grisoni, F. & Schneider, G. Geometric deep learning on molecular representations. Nat. Mach. Intell. 3, 1023–1032 (2021).Article

Atz，K.，Grisoni，F。＆Schneider，G。分子表征的几何深度学习。自然马赫数。因特尔。31023-1032（2021）。文章

Google Scholar

谷歌学者

Gainza, P. et al. Deciphering interaction fingerprints from protein molecular surfaces using geometric deep learning. Nat. Methods 17, 184–192 (2020).Article

Gainza，P。等人。使用几何深度学习从蛋白质分子表面破译相互作用指纹。自然方法17184-192（2020）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Tubiana, J., Schneidman-Duhovny, D. & Wolfson, H. J. ScanNet: an interpretable geometric deep learning model for structure-based protein binding site prediction. Nat. Methods 19, 730–739 (2022).Article

。自然方法19730-739（2022）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Gainza, P. et al. De novo design of protein interactions with learned surface fingerprints. Nature 617, 176–184 (2023).Article

Gainza，P.等人。从头设计具有学习表面指纹的蛋白质相互作用。。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Zhang, O. et al. Learning on topological surface and geometric structure for 3D molecular generation. Nat. Comput. Sci. 3, 849–859 (2023).Article

Zhang，O.等。学习拓扑表面和几何结构以生成三维分子。自然计算。科学。3849-859（2023）。文章

PubMed

PubMed

Google Scholar

谷歌学者

Sverrisson, F., Feydy, J., Correia, B. E. & Bronstein, M. M. Fast end-to-end learning on protein surfaces. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 15267–15276 (IEEE, 2021).Ektefaie, Y., Dasoulas, G., Noori, A., Farhat, M. & Zitnik, M. Multimodal learning with graphs.

Sverrisson，F.，Feydy，J.，Correia，B.E。和Bronstein，M.M。蛋白质表面的快速端到端学习。。Ektefaie，Y.，Dasoulas，G.，Noori，A.，Farhat，M。＆Zitnik，M。使用图形进行多模态学习。

Nat. Mach. Intell. 5, 340–350 (2023).Article .

纳特，去吧。英特尔。5, 340–350 (2023).第条。

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Igashov, I. et al. Decoding surface fingerprints for protein–ligand interactions. In ICLR2022 Machine Learning for Drug Discovery (2022).Li, S. et al. PocketAnchor: learning structure-based pocket representations for protein–ligand interaction prediction. Cell Syst. 14, 692–705.e696 (2023).Article .

Igashov，I。等人。解码蛋白质-配体相互作用的表面指纹。在ICLR2022药物发现的机器学习中（2022）。Li，S。等。PocketAnchor：基于学习结构的蛋白质-配体相互作用预测口袋表示。细胞系统。14692–705.e696（2023）。文章。

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Somnath, V. R., Bunne, C. & Krause, A. Multi-scale representation learning on proteins. In Advances in Neural Information Processing Systems 34, 25244–25255 (Curran Associates, 2021).Chen, L. et al. TransformerCPI: improving compound-protein interaction prediction by sequence-based deep learning with self-attention mechanism and label reversal experiments.

Somnath，V.R.，Bunne，C。和Krause，A。蛋白质的多尺度表征学习。《神经信息处理系统进展》3425244–25255（Curran Associates，2021）。Chen，L.等。TransformerCPI：通过基于序列的深度学习和自我注意机制以及标记逆转实验来改进复合蛋白质相互作用的预测。

Bioinformatics 36, 4406–4414 (2020).Article .

生物信息学364406-4414（2020）。文章。

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Chen, L. et al. Sequence-based drug design as a concept in computational drug design. Nat. Commun. 14, 4217 (2023).Article

Chen，L.等人。基于序列的药物设计作为计算药物设计的概念。国家公社。144217（2023）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Bryant, P., Kelkar, A., Guljas, A., Clementi, C. & Noé, F. Structure prediction of protein–ligand complexes from sequence information with Umol. Nat. Commun. 15, 4536 (2024).Article

Bryant，P.，Kelkar，A.，Guljas，A.，Clementi，C。＆Noé，F。从Umol的序列信息预测蛋白质-配体复合物的结构。国家公社。154536（2024）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Wang, R., Fang, X., Lu, Y. & Wang, S. The PDBbind database: collection of binding affinities for protein–ligand complexes with known three-dimensional structures. J. Med. Chem. 47, 2977–2980 (2004).Article

Wang，R.，Fang，X.，Lu，Y。＆Wang，S。PDBbind数据库：收集具有已知三维结构的蛋白质-配体复合物的结合亲和力。J、 医学化学。472977-2980（2004）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Hartshorn, M. J. et al. Diverse, high-quality test set for the validation of protein–ligand docking performance. J. Med. Chem. 50, 726–741 (2007).Article

Hartshorn，M.J.等人。用于验证蛋白质-配体对接性能的多种高质量测试集。J、 医学化学。50726-741（2007）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Bauer, M. R., Ibrahim, T. M., Vogel, S. M. & Boeckler, F. M. Evaluation and optimization of virtual screening workflows with DEKOIS 2.0–a public library of challenging docking benchmark sets. J. Chem. Inf. Model. 53, 1447–1462 (2013).Article

Bauer，M.R.，Ibrahim，T.M.，Vogel，S.M。和Boeckler，F.M。使用DEKOIS 2.0评估和优化虚拟筛选工作流程-一个具有挑战性的对接基准集公共图书馆。J、 化学。Inf.模型。531447-1462（2013）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Landrum, G. RDKit: Open-source Cheminformatics. (2006).Su, M., Feng, G., Liu, Z., Li, Y. & Wang, R. Tapping on the black box: How is the scoring power of a machine-learning scoring function dependent on the training set? J. Chem. Inf. Model. 60, 1122–1136 (2020).Article

Landrum，G。RDKit：开源化学信息学。（2006年）。Su，M.，Feng，G.，Liu，Z.，Li，Y.＆Wang，R.敲打黑匣子：机器学习评分函数的评分能力如何依赖于训练集？J、 化学。Inf.模型。601122-1136（2020）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Bax, B., Chung, C. W. & Edge, C. Getting the chemistry right: protonation, tautomers and the importance of H atoms in biological chemistry. Acta Crystallogr. D. Struct. Biol. 73, 131–140 (2017).Article

。晶体学报。D、 结构。生物学73131-140（2017）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Wang, Z. et al. Comprehensive evaluation of ten docking programs on a diverse set of protein–ligand complexes: the prediction accuracy of sampling power and scoring power. Phys. Chem. Chem. Phys. 18, 12964–12975 (2016).Article

Wang，Z.等人。对多种蛋白质-配体复合物的十种对接程序的综合评估：采样能力和评分能力的预测准确性。物理。化学。化学。物理。1812964-12975（2016）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Su, M. et al. Comparative assessment of scoring functions: the CASF-2016 update. J. Chem. Inf. Model. 59, 895–913 (2019).Article

Su，M.等人。评分功能的比较评估：CASF-2016更新。J、 化学。Inf.模型。59895-913（2019）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Adeshina, Y. O., Deeds, E. J. & Karanicolas, J. Machine learning classification can reduce false positives in structure-based virtual screening. Proc. Natl Acad. Sci. USA 117, 18477–18488 (2020).Article

机器学习分类可以减少基于结构的虚拟筛选中的误报。程序。国家科学院。科学。美国11718477–18488（2020）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Zhang, X. et al. TocoDecoy: a new approach to design unbiased datasets for training and benchmarking machine-learning scoring functions. J. Med. Chem. 65, 7918–7932 (2022).Article

Zhang，X。et al。TocoDecoy：一种设计用于训练和基准机器学习评分函数的无偏数据集的新方法。J、 医学化学。657918-7932（2022）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Dong, T., Yang, Z., Zhou, J. & Chen, C. Y.-C. Equivariant flexible modeling of the protein–ligand binding pose with geometric deep learning. J. Chem. Theory Comput. 19, 8446–8459 (2023).Article

Dong，T.，Yang，Z.，Zhou，J.＆Chen，C.Y.-C。通过几何深度学习对蛋白质-配体结合姿势进行等变柔性建模。J、 化学。理论计算。198446-8459（2023）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Huang, Y. et al. Re-dock: towards flexible and realistic molecular docking with diffusion bridge. In International Conference on Machine Learning 235, 20474–20489 (PMLR, 2024).Feng, Z. et al. Targeting colorectal cancer with small-molecule inhibitors of ALDH1B1. Nat. Chem. Biol. 18, 1065–1075 (2022).Article .

Huang，Y.等。Re dock：通过扩散桥实现灵活而现实的分子对接。在国际机器学习会议23520474-20489（PMLR，2024）中。Feng，Z.等人。用ALDH1B1的小分子抑制剂靶向结直肠癌。自然化学。生物学181065-1075（2022）。文章。

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

O’Boyle, N. M. et al. Open Babel: an open chemical toolbox. J. Cheminform. 3, 33 (2011).Article

O'Boyle，N.M.等人，《打开巴贝尔：一个开放的化学工具箱》。J、 化学信息。3，33（2011）。文章

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Word, J. M., Lovell, S. C., Richardson, J. S. & Richardson, D. C. Asparagine and glutamine: using hydrogen atom contacts in the choice of side-chain amide orientation. J. Mol. Biol. 285, 1735–1747 (1999).Article

Word，J.M.，Lovell，S.C.，Richardson，J.S。＆Richardson，D.C。天冬酰胺和谷氨酰胺：在选择侧链酰胺取向时使用氢原子接触。J、 分子生物学。2851735-1747（1999）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Batzner, S. et al. E(3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials. Nat. Commun. 13, 2453 (2022).Article

Batzner，S。等人，E（3）-等变图神经网络，用于数据高效和准确的原子间势。国家公社。。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Jing, B., Corso, G., Chang, J., Barzilay, R. & Jaakkola, T. S. Torsional diffusion for molecular conformer generation. In Advances in Neural Information Processing Systems 35, 24240–24253 (Curran Associates, 2022).Sanner, M. F., Olson, A. J. & Spehner, J. C. Reduced surface: an efficient way to compute molecular surfaces.

Jing，B.，Corso，G.，Chang，J.，Barzilay，R。＆Jaakkola，T.S。分子构象异构体生成的扭转扩散。《神经信息处理系统进展》3524240–24253（Curran Associates，2022）。Sanner，M.F.，Olson，A.J。＆Spehner，J.C。还原表面：计算分子表面的有效方法。

Biopolymers 38, 305–320 (1996).Article .

生物聚合物38305-320（1996）。文章。

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Zhou, Q. Pymesh—geometry processing library for Python. https://github.com/PyMesh/PyMesh (2019).Geiger, M. & Smidt, T. e3nn: Euclidean neural networks. Preprint at https://arxiv.org/abs/2207.09453 (2022).Bishop, C. M. Mixture Density Networks. (1994).Loshchilov, I. & Hutter, F. Decoupled weight decay regularization.

Zhou，Q。Python的Pymesh几何处理库。https://github.com/PyMesh/PyMesh。Geiger，M。＆Smidt，T。e3nn：欧几里得神经网络。预印于https://arxiv.org/abs/2207.09453（2022年）。Bishop，C.M。混合密度网络。。Loshchilov，I。＆Hutter，F。解耦权重衰减正则化。

In International Conference on Learning Representations (2017).Eastman, P. et al. OpenMM 7: rapid development of high performance algorithms for molecular dynamics. PLoS Comput. Biol. 13, e1005659 (2017).Article .

在国际学习表征会议（2017年）上。Eastman，P。等人。OpenMM 7：分子动力学高性能算法的快速发展。PLoS计算机。生物学杂志13，e1005659（2017）。文章。

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Maier, J. A. et al. Ff14SB: improving the accuracy of protein side chain and backbone parameters from ff99SB. J. Chem. Theory Comput. 11, 3696–3713 (2015).Article

Maier，J.A.等人，Ff14SB：提高ff99SB蛋白质侧链和骨架参数的准确性。J、 化学。理论计算。113696-3713（2015）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Boothroyd, S. et al. Development and benchmarking of Open Force Field 2.0.0: the Sage small molecule force field. J. Chem. Theory Comput. 19, 3251–3275 (2023).Article

Boothroyd，S.等人。开放力场2.0.0的开发和基准测试：Sage小分子力场。J、 化学。理论计算。193251–3275（2023）。文章

CAS

中科院

PubMed

PubMed

PubMed Central

公共医学中心

Google Scholar

谷歌学者

Wang, J., Wolf, R. M., Caldwell, J. W., Kollman, P. A. & Case, D. A. Development and testing of a general amber force field. J. Comput. Chem. 25, 1157–1174 (2004).Article

Wang，J.，Wolf，R.M.，Caldwell，J.W.，Kollman，P.A。和Case，D.A。通用琥珀力场的开发和测试。J、 计算机。化学。251157-1174（2004）。文章

CAS

中科院

PubMed

PubMed

Google Scholar

谷歌学者

Cao, D. Code for paper ‘SurfDock is a surface-informed diffusion generative model for reliable and accurate protein–ligand complex prediction’ (v0.0.1). Zenodo https://doi.org/10.5281/zenodo.13933663 (2024).Download referencesAcknowledgementsWe thank Y. Yu, S. Lu and H. Zheng for providing the original docking results of Uni-Dock, and R.

Cao，D。论文代码“SurfDock是一种表面知情的扩散生成模型，可用于可靠而准确的蛋白质-配体复合物预测”（v0.0.1）。泽诺多https://doi.org/10.5281/zenodo.13933663（2024年）。下载参考文献致谢我们感谢Y.Yu，S.Lu和H.Zheng提供Uni Dock和R的原始对接结果。

Li, X. He and V. Ram Somnath for their valuable feedback and insightful discussions. This work was supported by National Natural Science Foundation of China (T2225002 and 82273855 to M.Z., 82204278 to X.L.), Strategic Priority Research Program of the Chinese Academy of sciences (XDB0850000), National Key Research and Development Program of China (2022YFC3400504 and 2023YFC2305904 to M.Z.), SIMM-SHUTCM Traditional Chinese Medicine Innovation Joint Research Program (E2G805H), The Youth Innovation Promotion Association CAS (2023296 to S.Z.), and Shanghai Municipal Science and Technology Major Project.

Li，X。He和V。Ram Somnath的宝贵反馈和深刻的讨论。这项工作得到了国家自然科学基金（T2225002和82273855至M.Z.，82204278至X.L.），中国科学院战略重点研究计划（XDB0850000），国家重点研究发展计划（2022YFC3400504和2023YFC2305904至M.Z.），SIMM-SHUTCM中医药创新联合研究计划（E2G805H），中国科学院青年创新促进会（2023296至S.Z.）和上海市科技重大项目的支持。

We thank the staff members of the Large-scale Protein Preparation System at the National Facility for Protein Science in Shanghai, Shanghai Advanced Research Institute, Chinese Academy of Science, China for providing technical support and assistance in data collection and analysis.Author informationAuthor notesThese authors contributed equally: Duanhua Cao, Mingan Chen, Runze Zhang.Authors and AffiliationsInnovation Institute for Artificial Intelligence in Medicine of Zhejiang University, College of Pharmaceutical Sciences, Zhejiang University, Hangzhou, ChinaDuanhua Cao & Mingyue ZhengDrug Discovery and Design Center, State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai, ChinaDuanhua Cao, Mingan Chen, Runze Zhang, Zhaokun Wang, Manlin Huang, Jie Yu, Xinyu Jiang, Zh.

我们感谢中国科学院上海高级研究所上海国家蛋白质科学研究所大规模蛋白质制备系统的工作人员在数据收集和分析方面提供的技术支持和帮助。作者信息作者注意到这些作者做出了同样的贡献：曹端华，陈明安，张润泽。作者和单位浙江大学医学人工智能创新研究所，浙江大学药学院，杭州，中国曹端华和郑明月药物发现与设计中心，药物研究国家重点实验室，中国科学院上海药物研究所，上海，曹端华，陈明安，张润泽，王兆坤，黄曼林，于杰，蒋新宇，Zh。

PubMed Google ScholarMingan ChenView author publicationsYou can also search for this author in

PubMed Google Scholarmaringan ChenView作者出版物您也可以在

PubMed Google ScholarRunze ZhangView author publicationsYou can also search for this author in

PubMed Google ScholarRunze ZhangView作者出版物您也可以在

PubMed Google ScholarZhaokun WangView author publicationsYou can also search for this author in

PubMed Google ScholarZhaokun WangView作者出版物您也可以在

PubMed Google ScholarManlin HuangView author publicationsYou can also search for this author in

PubMed Google ScholarManlin HuangView作者出版物您也可以在

PubMed Google ScholarJie YuView author publicationsYou can also search for this author in

PubMed Google ScholarJie YuView作者出版物您也可以在

PubMed Google ScholarXinyu JiangView author publicationsYou can also search for this author in

PubMed谷歌学者Xinyu JiangView作者出版物您也可以在

PubMed Google ScholarZhehuan FanView author publicationsYou can also search for this author in

PubMed Google ScholarWei ZhangView author publicationsYou can also search for this author in

PubMed Google ScholarWei ZhangView作者出版物您也可以在

PubMed Google ScholarHao ZhouView author publicationsYou can also search for this author in

PubMed Google ScholarHao ZhouView作者出版物您也可以在

PubMed Google ScholarXutong LiView author publicationsYou can also search for this author in

PubMed Google ScholarXutong LiView作者出版物您也可以在

PubMed Google ScholarZunyun FuView author publicationsYou can also search for this author in

PubMed Google ScholarSulin ZhangView author publicationsYou can also search for this author in

PubMed Google ScholarSulin ZhangView作者出版物您也可以在

PubMed Google ScholarMingyue ZhengView author publicationsYou can also search for this author in

PubMed Google ScholarmamingYue ZhengView作者出版物您也可以在

PubMed Google ScholarContributionsD.C., M.C. and R.Z. contributed equally. M.Z. conceived the research project. D.C. developed the primary method and code. M.C. developed the force field optimize code. Z.W. and M.H. contributed to the biological experiments on ALDH1B1 inhibitors.

PubMed谷歌学术贡献SD。C、 ，M.C.和R.Z.的贡献相同。M、 。D、 C.开发了主要方法和代码。M、 C.开发了力场优化代码。Z、 W.和M.H.为ALDH1B1抑制剂的生物学实验做出了贡献。

R.Z., X.J. and J.Y. assisted in analyzing the main baselines in the paper. All authors contributed to the analysis of the results. D.C., M.C. and M.Z. wrote the paper. All authors read and approved the manuscript.Corresponding authorCorrespondence to.

R、 Z.，X.J.和J.Y.协助分析了论文中的主要基线。所有作者都为结果分析做出了贡献。D、 C.，M.C.和M.Z.写了这篇论文。所有作者都阅读并批准了手稿。对应作者对应。

Mingyue Zheng.Ethics declarations

郑明月。道德宣言

Competing interests

相互竞争的利益

The authors declare no competing interests.

作者声明没有利益冲突。

Peer review

同行评审

Peer review information

同行评审信息

Nature Methods thanks Mayukh Chakrabarti, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editor: Arunima Singh, in collaboration with the Nature Methods team.

Nature Methods感谢Mayukh Chakrabarti和另一位匿名审稿人对这项工作的同行评审做出的贡献。主要处理编辑：Arunima Singh，与Nature Methods团队合作。

Additional informationPublisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Extended dataExtended Data Fig. 1 The Performance across Different Docking Methods.(For rotatable bonds: rigid: ≤ 5; medium: 5-10; flexible: > 10) a: Impact of rotatable bonds on docking accuracy on PDBbind time-split set, and the distribution of RMSD.

Additional informationPublisher的注释Springer Nature在已发布的地图和机构隶属关系中的管辖权主张方面保持中立。扩展数据扩展数据图1不同对接方法的性能。（对于可旋转键：刚性：≤5；中等：5-10；灵活：＞10）a：可旋转键对PDBbind时间分裂集对接精度的影响，以及RMSD的分布。

b: Impact of rotatable bonds on docking accuracy on PoseBusters set, and the distribution of RMSD.Source dataExtended Data Fig. 2 Cases of Ligand Poses Predicted by Different Docking Methods in Comparison with Crystal Ligand Pose.The crystal ligand poses are in green, while the predicted ligand poses are in other colors.

b： 旋转键对PoseBusters集对接精度的影响，以及RMSD的分布。源数据扩展数据图2与晶体配体姿势相比，通过不同对接方法预测的配体姿势的情况。。

a, b, c: cases with different ligand complexity where SurfDock outperforms traditional methods. d, e: cases where crystal ligand pose has clashes with the protein pocket, but SurfDock predictions avoid these clashes.Extended Data Table 1 Comparative Analysis of Docking Performances on DockGen datasetFull size tableExtended Data Table 2 Apo Docking Performances of Different Methods on PDBbind2020Full size tableSupplementary informationSupplementary InformationSupplementary Figs.

a、 b，c：具有不同配体复杂性的情况，其中SurfDock优于传统方法。d、 e：晶体配体姿势与蛋白质口袋发生冲突的情况，但SurfDock预测避免了这些冲突。扩展数据表1 DockGen数据集对接性能的比较分析全尺寸表扩展数据表2 PDBBIND2020上不同方法的Apo对接性能全尺寸表补充信息补充信息补充图。

1–8, Supplementary Discussion and Supplementary Tables 1–3.Reporting SummarySupplementary Data 1Statistical Source Data for Supplementary Fig. 1.Supplementary Data 2Statistical Source Data for Supplementary Fig. 2.Supplementary Data 3Statistical Source Data for Supplementary Fig. 3.Supplementary Data 4Statistical Source Data for Supplementary Fig.

1-8，补充讨论和补充表1-3。报告摘要补充数据1补充图1的统计源数据。补充数据2补充图2的统计源数据。补充数据3补充图3的统计源数据。补充数据4补充图的统计源数据。

4.Supplementary Data 5Statistical Source Data for Supplementary Fig. 5.Supplementary Data 7Statistical Source Data for Supplementary Fig. 7.Supplementary Data 8Statistical Source Data for Supplementary Fig. 8.Source dataSource Data Fi.

4。补充数据5补充图5的统计源数据。补充数据7补充图7的统计源数据。补充数据8补充图8的统计源数据。源数据源数据Fi。

Nat Methods (2024). https://doi.org/10.1038/s41592-024-02516-yDownload citationReceived: 19 January 2024Accepted: 16 October 2024Published: 27 November 2024DOI: https://doi.org/10.1038/s41592-024-02516-yShare this articleAnyone you share the following link with will be able to read this content:Get shareable linkSorry, a shareable link is not currently available for this article.Copy to clipboard.

Nat方法（2024）。https://doi.org/10.1038/s41592-024-02516-yDownload引文接收日期：2024年1月19日接受日期：2024年10月16日发布日期：2024年11月27日OI：https://doi.org/10.1038/s41592-024-02516-yShare本文与您共享以下链接的任何人都可以阅读此内容：获取可共享链接对不起，本文目前没有可共享的链接。复制到剪贴板。

Provided by the Springer Nature SharedIt content-sharing initiative

由Springer Nature SharedIt内容共享计划提供