CoChat
Sign Up
PreprintCited by 274Protein Structure and DynamicsMachine Learning in BioinformaticsComputational Drug Discovery Methods

The DOI matches its Crossref record; 1 of the 1 references it cites were found in a registry or an index.

ScanNet: An interpretable geometric deep learning model for structure-based protein binding site prediction

Open access
Tubiana, Jérôme, Schneidman-Duhovny, Dina, Wolfson, Haim J.·2021·Nature Methods
DOI 10.1101/2021.09.05.459013

Preprint — published as 10.1038/s41592-022-01490-7 · Paper page

Open this paper in CoChat — free account

Abstract

Predicting the functional sites of a protein from its structure, such as the binding sites of small molecules, other proteins or antibodies sheds light on its function in vivo . Currently, two classes of methods prevail: Machine Learning (ML) models built on top of handcrafted features and comparative modeling. They are respectively limited by the expressivity of the handcrafted features and the availability of similar proteins. Here, we introduce ScanNet, an end-to-end, interpretable geometric deep learning model that learns features directly from 3D structures. ScanNet builds representations of atoms and amino acids based on the spatio-chemical arrangement of their neighbors. We train ScanNet for detecting protein-protein and protein-antibody binding sites, demonstrate its accuracy - including for unseen protein folds - and interpret the filters learned. Finally, we predict epitopes of the SARS-CoV-2 spike protein, validating known antigenic regions and predicting previously uncharacterized ones. Overall, ScanNet is a versatile, powerful, and interpretable model suitable for functional site prediction tasks. A webserver for ScanNet is available from http://bioinfo3d.cs.tau.ac.il/ScanNet/

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1101/2021.09.05.459013

Cites

1 of 1 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Flagged first, then by citations

Research works printed but not linked: 98

  1. 10.1126/science.1251652 · DOI · Paper page
  2. K. Tunyasuvunakool , J. Adler , Z. Wu , T. Green , M. Zielinski , A. Žídek , A. Bridgland , A. Cowie , C. Meyer , A. Laydon , et al., Highly accurate protein structure prediction for the human proteome, Nature, 1 (2021). · DOI · Paper page
  3. 10.1016/j.sbi.2010.08.001 · DOI
  4. 10.1093/nar/gki482 · DOI
  5. 10.1021/ci100265x · DOI
  6. 10.1073/pnas.1005894107 · DOI · Paper page
  7. 10.1186/1471-2105-12-1 · DOI
  8. 10.1093/nar/gkr997 · DOI
  9. 10.1186/1471-2105-13-1 · DOI
  10. 10.1016/j.febslet.2015.10.003 · DOI
  11. 10.1093/bib/bbv027 · DOI
  12. 10.1016/j.jmb.2004.02.040 · DOI
  13. 2006, Exploiting sequence and structure homologs to identify protein–protein binding sites, Proteins: Structure, Function, and Bioinformatics, 62, 630, 10.1002/prot.20741 · DOI
  14. 2007, Prediction-based fingerprints of protein–protein interactions, Proteins: Structure, Function, and Bioinformatics, 66, 630, 10.1002/prot.21248 · DOI
  15. 10.1093/bioinformatics/btn199 · DOI
  16. 2019, Coupling dynamics and evolutionary information with structure to identify protein regulatory and functional binding sites, Proteins: Structure, Function, and Bioinformatics, 87, 850, 10.1002/prot.25749 · DOI
  17. 10.1016/0968-0004(87)90231-3 · DOI
  18. 10.1006/jmbi.1998.1843 · DOI · Paper page
  19. 2021, A lysine–cysteine redox switch with an nos bridge regulates enzyme function, Nature, 593, 460, 10.1038/s41586-021-03513-3 · DOI
  20. A. Elnaggar , M. Heinzinger , C. Dallago , G. Rihawi , Y. Wang , L. Jones , T. Gibbs , T. Feher , C. Angerer , M. Steinegger , et al., Prottrans: towards cracking the language of life’s code through self-supervised deep learning and high performance computing, arXiv preprint arXiv:2007.06225 (2020). · DOI
  21. A. Rives , J. Meier , T. Sercu , S. Goyal , Z. Lin , J. Liu , D. Guo , M. Ott , C. L. Zitnick , J. Ma , et al., Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences, Proceedings of the National Academy of Sciences 118(2021). · DOI · Paper page
  22. J. Ingraham , A. Riesselman , C. Sander , and D. Marks , Learning protein structure with a differentiable simulator, in International Conference on Learning Representations (2018).
  23. X. Jing and J. Xu , Fast and effective protein model refinement by deep graph neural networks, bioRxiv (2020). · DOI
  24. 2021, Graphqa: protein model quality assessment using graph convolutional networks, Bioinformatics, 37, 360, 10.1093/bioinformatics/btaa714 · DOI
  25. I. Wallach , M. Dzamba , and A. Heifets , Atomnet: a deep convolutional neural network for bioactivity prediction in structure-based drug discovery, arXiv preprint arXiv:1510.02855 (2015). · arXiv
  26. 10.1021/acs.jcim.6b00740 · DOI
  27. 2019, Protein model quality assessment using 3d oriented convolutional neural networks, Bioinformatics, 35, 3313, 10.1093/bioinformatics/btz122 · DOI
  28. 2019, End-to-end learning on 3d protein structure for interface prediction, Advances in Neural Information Processing Systems, 32, 15642
  29. 2020, Protein docking model evaluation by 3d deep convolutional neural networks, Bioinformatics, 36, 2113, 10.1093/bioinformatics/btz870 · DOI
  30. I. Igashov , K. Olechnovic , M. Kadukova , Č. Venclovas , and S. Grudinin , Vorocnn: Deep convolutional neural network built on 3d voronoi tessellation of protein structures, bioRxiv (2020). · DOI
  31. N. Renaud , C. Geng , S. Georgievska , F. Ambrosetti , L. Ridder , D. F. Marzella , A. M. Bonvin , and L. C. Xue , Deeprank: A deep learning framework for data mining 3d protein-protein interfaces, Biorxiv (2021). · DOI
  32. S. Eismann , P. Suriana , B. Jing , R. J. Townshend , and R. O. Dror , Protein model quality assessment using rotation-equivariant, hierarchical neural networks, arXiv preprint arXiv:2011.13557 (2020). · arXiv
  33. 2020, Deciphering interaction fingerprints from protein molecular surfaces using geometric deep learning, Nature Methods, 17, 184, 10.1038/s41592-019-0666-6 · DOI
  34. F. Sverrisson , J. Feydy , B. E. Correia , and M. M. Bronstein , Fast end-to-end learning on protein surfaces, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2021) pp. 15272–15281. · DOI
  35. 10.1109/MSP.2017.2693418 · DOI · Paper page
  36. M. M. Bronstein , J. Bruna , T. Cohen , and P. Veličković , Geometric deep learning: Grids, groups, graphs, geodesics, and gauges, arXiv preprint arXiv:2104.13478 (2021). · arXiv
  37. J. Gilmer , S. S. Schoenholz , P. F. Riley , O. Vinyals , and G. E. Dahl , Neural message passing for quantum chemistry, in International conference on machine learning (PMLR, 2017) pp. 1263–1272.
  38. P. Veličković , G. Cucurull , A. Casanova , A. Romero , P. Lio , and Y. Bengio , Graph attention networks, arXiv preprint arXiv:1710.10903 (2017). · arXiv
  39. 10.1016/j.jmb.2004.10.077 · DOI · Paper page
  40. 2007, Protein–protein interaction hotspots carved into sequences, PLoS computational biology, 3, e119, 10.1371/journal.pcbi.0030119 · DOI
  41. 2018, Pdb-wide identification of biological assemblies from conserved quaternary structure geometry, Nature methods, 15, 67, 10.1038/nmeth.4510 · DOI
  42. 2018, Dockground: a comprehensive data resource for modeling of protein complexes, Protein Science, 27, 172, 10.1002/pro.3295 · DOI
  43. T. Chen and C. Guestrin , Xgboost: A scalable tree boosting system, in Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining (2016) pp. 785–794. · DOI · Paper page
  44. M. Shatsky , R. Nussinov , and H. J. Wolfson , Multiprot—a multiple protein structural alignment algorithm, in International Workshop on Algorithms in Bioinformatics (Springer, 2002) pp. 235–250. · DOI
  45. 2021, Cath: increased structural coverage of functional space, Nucleic acids research, 49, D266, 10.1093/nar/gkaa1079 · DOI · Paper page
  46. 10.1002/pro.3280 · DOI
  47. 10.1093/nar/gkt1043 · DOI · Paper page
  48. 10.1371/journal.pcbi.1002829 · DOI · Paper page
  49. M. Yuan , D. Huang , C.-C. D. Lee , N. C. Wu , A. M. Jackson , X. Zhu , H. Liu , L. Peng , M. J. van Gils , R. W. Sanders , et al., Structural and functional ramifications of antigenic drift in recent sars-cov-2 variants, Science (2021). · DOI · Paper page
  50. M. M. Sauer , M. A. Tortorici , Y.-J. Park , A. C. Walls , L. Homad , O. J. Acton , J. E. Bowen , C. Wang , X. Xiong , W. de van der Schueren , et al., Structural basis for broad coronavirus neutralization, Nature Structural & Molecular Biology, 1 (2021). · DOI
  51. 10.1126/science.abb9983 · DOI
  52. R. Evans , M. O’Neill , A. Pritzel , N. Antropova , A. W. Senior , T. Green , A. Žídek , R. Bates , S. Blackwell , J. Yim , et al., Protein complex prediction with alphafold-multimer, Biorxiv (2021). · DOI · Paper page
  53. 10.1021/bi00196a004 · DOI
  54. 2007, Gaba production by glutamic acid decarboxylase is regulated by a dynamic catalytic loop, Nature structural & molecular biology, 14, 280, 10.1038/nsmb1228 · DOI
  55. 10.1002/pro.3235 · DOI · Paper page
  56. L. Van der Maaten and G. Hinton , Visualizing data using t-sne., Journal of machine learning research 9(2008).
  57. 1983, Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features, Biopolymers: Original Research on Biomolecules, 22, 2577, 10.1002/bip.360221211 · DOI · Paper page
  58. R. Amaro and A. Mulholland , Biomolecular simulations in the time of covid19, and after, Computing in Science & Engineering (2020). · DOI
  59. 10.1093/bioinformatics/btp163 · DOI · Paper page
  60. 2012, Hhblits: lightning-fast iterative protein sequence searching by hmm-hmm alignment, Nature methods, 9, 173, 10.1038/nmeth.1818 · DOI · Paper page
  61. 10.1093/nar/gkw1081 · DOI
  62. 2018, Inverse statistical physics of protein sequences: a key issues review, Reports on Progress in Physics, 81, 032601, 10.1088/1361-6633/aa9965 · DOI
  63. L. Posani , Inference and modeling of biological networks: a statistical-physics approach to neural attractors and protein fitness landscapes, Ph.D. thesis, Université Paris sciences et lettres (2018).
  64. W. Chen , X. Han , G. Li , C. Chen , J. Xing , Y. Zhao , and H. Li , Deep rbfnet: Point cloud feature learning using radial basis functions, arXiv preprint arXiv:1812.04302 (2018). · arXiv
  65. C. R. Qi , H. Su , K. Mo , and L. J. Guibas , Pointnet: Deep learning on point sets for 3d classification and segmentation, in Proceedings of the IEEE conference on computer vision and pattern recognition (2017) pp. 652–660.
  66. C. R. Qi , L. Yi , H. Su , and L. J. Guibas , Pointnet++: Deep hierarchical feature learning on point sets in a metric space, arXiv preprint arXiv:1706.02413 (2017). · arXiv
  67. I. Igashov , N. Pavlichenko , and S. Grudinin , Spherical convolutions on molecular graphs for protein model quality assessment, Machine Learning: Science and Technology (2021). · DOI
  68. 2019, Learning protein constitutive motifs from sequence data, Elife, 8, e39397, 10.7554/eLife.39397 · DOI
  69. S. Ioffe and C. Szegedy , Batch normalization: Accelerating deep network training by reducing internal covariate shift, in International conference on machine learning (PMLR, 2015) pp. 448–456.
  70. 2011, Scikit-learn: Machine learning in python, the Journal of machine Learning research, 12, 2825
  71. J. Long , E. Shelhamer , and T. Darrell , Fully convolutional networks for semantic segmentation, in Proceedings of the IEEE conference on computer vision and pattern recognition (2015) pp. 3431–3440. · DOI · Paper page
  72. D. P. Kingma and J. Ba , Adam: A method for stochastic optimization, arXiv preprint arXiv:1412.6980 (2014). · arXiv
  73. M. Abadi , P. Barham , J. Chen , Z. Chen , A. Davis , J. Dean , M. Devin , S. Ghemawat , G. Irving , M. Isard , et al., Tensorflow: A system for large-scale machine learning, in 12th {USENIX} symposium on operating systems design and implementation ({OSDI} 16) (2016) pp. 265–283.
  74. 10.1093/bioinformatics/btn222 · DOI
  75. 10.1016/S0969-2126(99)80097-5 · DOI
  76. 10.1002/(SICI)1097-0282(199603)38:3<305::AID-BIP4>3.0.CO;2-Y · DOI
  77. 10.1002/bip.360250705 · DOI
  78. 10.1093/bioinformatics/bts565 · DOI · Paper page
  79. 10.1093/bioinformatics/btl158 · DOI
  80. 10.1093/bioinformatics/bty121 · DOI
  81. 10.1186/s12859-014-0430-y · DOI
  82. U. Ghani , I. Desta , A. Jindal , O. Khan , G. Jones , S. Kotelnikov , D. Padhorny , S. Vajda , and D. Kozakov , Improved docking of protein models by a combination of alphafold2 and cluspro, BioRxiv (2021). · DOI · Paper page
  83. M. Mirdita , K. Schütze , Y. Moriwaki , L. Heo , S. Ovchinnikov , and M. Steinegger , Colabfold-making protein folding accessible to all, (2021). · DOI · Paper page
  84. 2019, Skempi 2.0: an updated benchmark of changes in protein–protein binding energy, kinetics and thermodynamics upon mutation, Bioinformatics, 35, 462, 10.1093/bioinformatics/bty635 · DOI · Paper page
  85. 10.1016/j.jmb.2015.07.016 · DOI · Paper page
  86. 2017, The rosetta all-atom energy function for macromolecular modeling and design, Journal of chemical theory and computation, 13, 3031, 10.1021/acs.jctc.7b00125 · DOI · Paper page
  87. 10.1093/bioinformatics/btq007 · DOI
  88. M. Yuan , H. Liu , N. C. Wu , C.-C. D. Lee , X. Zhu , F. Zhao , D. Huang , W. Yu , Y. Hua , H. Tien , et al., Structural basis of a public antibody response to sars-cov-2, BioRxiv (2020). · DOI
  89. 2020, An alternative binding mode of ighv3-53 antibodies to the sars-cov-2 receptor binding domain, Cell reports, 33, 108274, 10.1016/j.celrep.2020.108274 · DOI
  90. 10.1016/j.cell.2020.09.049 · DOI · Paper page
  91. 10.1126/science.abd0827 · DOI · Paper page
  92. 2021, A combination of cross-neutralizing antibodies synergizes to prevent sars-cov-2 and sars-cov pseudovirus infection, Cell host & microbe, 29, 806, 10.1016/j.chom.2021.04.005 · DOI
  93. 2020, Cross-neutralization of a sars-cov-2 antibody to a functionally conserved site is mediated by avidity, Immunity, 53, 1272, 10.1016/j.immuni.2020.10.023 · DOI
  94. 10.1016/j.chom.2021.03.005 · DOI · Paper page
  95. 2014, Receptor mimicry by antibody f045–092 facilitates universal binding to the h3 subtype of influenza virus, Nature communications, 5, 1
  96. 10.1126/science.1093155 · DOI
  97. 10.1016/j.cell.2016.04.010 · DOI
  98. 10.1126/science.abd7331 · DOI · Paper page

Metadata

Authors
Tubiana, J., Schneidman-Duhovny, D., & Wolfson, H. J.
Affiliations
Tel Aviv University (IL)
Corresponding author
  • Tubiana, Jérôme
  • Wolfson, Haim J.
ORCID
Venue
Nature Methods
Publisher
openRxiv
Published
2021-09-06
Volume
19
Issue
6
Pages
730-739
ISSN
1548-7091 (ISSN-L)
DOI
10.1101/2021.09.05.459013
OpenAlex
W4229031603
Cited by
274
References
1
Citations by year
2026: 70 · 2025: 75 · 2024: 69 · 2023: 48 · 2022: 11 · 2021: 1
FWCI
16.81
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • Nature Methods (publishedVersion)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by)
  • Research Square (acceptedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
  • Blavatnik Family Foundation
  • Human Frontier Science Program (LT001058/2019-C)
  • Hebrew University of Jerusalem
  • Israel Science Foundation (ISF 1466/18, 1466/18)
  • Tel Aviv University
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
  • Life Sciences › Pharmacology, Toxicology and Pharmaceutics › Drug Discovery › Computational Drug Discovery Methods
Keywords
protein binding site prediction, geometric deep learning, model interpretability, protein structure, protein-protein binding, epitope prediction, SARS-CoV-2 spike protein
MeSH terms
Deep Learning, COVID-19, SARS-CoV-2, Binding Sites, Humans, Protein Binding, Proteins, Spike Glycoprotein, Coronavirus

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review