CoChat
Sign Up
PreprintCited by 108Computational Drug Discovery MethodsMachine Learning in BioinformaticsProtein Interaction Studies and Fluorescence Analysis

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

PSICHIC: physicochemical graph neural network for learning protein-ligand interaction fingerprints from sequence data

Open access
Koh, Huan Yee, Nguyen, Anh T.N., Pan, Shirui, May, Lauren T., Webb, Geoffrey I.·2023·Nature Machine Intelligence
DOI 10.1101/2023.09.17.558145

Preprint — published as 10.1038/s42256-024-00847-1

Open this paper in CoChat — free accountOpen PDF

Abstract

In drug discovery, determining the binding affinity and functional effects of small-molecule ligands on proteins is critical. Current computational methods can predict these protein-ligand interaction properties but often lose accuracy without high-resolution protein structures and falter in predicting functional effects. We introduce PSICHIC (PhySIcoCHemICal graph neural network), a framework uniquely incorporating physicochemical constraints to decode interaction fingerprints directly from sequence data alone. This enables PSICHIC to attain first-of-its-kind emergent capabilities in deciphering mechanisms underlying protein-ligand interactions, achieving state-of-the-art accuracy and interpretability. Trained on identical protein-ligand pairs without structural data, PSICHIC matched and even surpassed leading structure-based methods in binding affinity prediction. In a library screening for adenosine A1 receptor agonists, PSICHIC discerned functional effects effectively, ranking the sole novel agonist within the top three. PSICHIC’s interpretable fingerprints identified protein residues and ligand atoms involved in interactions. We foresee PSICHIC reshaping virtual screening and deepening our understanding of protein-ligand interactions.

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 · 10.1101/2023.09.17.558145

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: 66

  1. 10.1038/nrd1549 · DOI
  2. 10.1038/462167a · DOI
  3. 10.1093/bioinformatics/btaa524 · DOI · Paper page
  4. 2020, Drug–target affinity prediction using graph neural network and contact maps, RSC Adv, 10, 20701, 10.1039/D0RA02297G · DOI · Paper page
  5. 2021, Machine learning approaches and databases for prediction of drug–target interaction: a survey paper, Briefings Bioinforma, 22, 247, 10.1093/bib/bbz157 · DOI
  6. Li, S. et al. Structure-aware interactive graph neural networks for the prediction of protein-ligand binding affinity. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, 975–985 (2021). · DOI
  7. 2022, Artificial intelligence in the prediction of protein–ligand interactions: recent advances and future directions, Briefings Bioinforma, 23, bbab476, 10.1093/bib/bbab476 · DOI
  8. 2022, Tankbind: Trigonometry-aware neural networks for drug-protein binding structure prediction, Adv. Neural Inf. Process. Syst, 35, 7236
  9. Bai, P. , Miljković, F. , John, B. & Lu, H. Interpretable bilinear attention network with domain adaptation improves drug–target prediction. Nat. Mach. Intell. 1–11 (2023). · DOI
  10. Ng, H. W. et al. Competitive molecular docking approach for predicting estrogen receptor subtype α agonists and antagonists. In BMC Bioinformatics, 1–15 (2014). · DOI
  11. 2015, Molecular docking screening using agonist-bound gpcr structures: probing the a2a adenosine receptor, J. Chem. Inf. Model, 55, 550, 10.1021/ci500639g · DOI
  12. 10.1021/acs.jcim.5b00066 · DOI
  13. 2022, Deepreal: a deep learning powered multi-scale modeling framework for predicting out-of-distribution ligand-induced gpcr activity, Bioinformatics, 38, 2561, 10.1093/bioinformatics/btac154 · DOI
  14. 2019, Pconsc4: fast, accurate and hassle-free contact predictions, Bioinformatics, 35, 2677, 10.1093/bioinformatics/bty1036 · DOI
  15. Rao, R. , Meier, J. , Sercu, T. , Ovchinnikov, S. & Rives, A. Transformer protein language models are unsupervised structure learners. In International Conference on Learning Representations (2021). · DOI · Paper page
  16. 10.1126/science.ade2574 · DOI · Paper page
  17. Jiang, M. et al. Sequence-based drug-target affinity prediction using weighted graph neural networks. BMC Genomics 23 (2022). · DOI
  18. 2022, Structure-aware multimodal deep learning for drug–protein interaction prediction, J. Chem. Inf. Model, 62, 1308, 10.1021/acs.jcim.2c00060 · DOI
  19. 2020, Deciphering interaction fingerprints from protein molecular surfaces using geometric deep learning, Nat. Methods, 17, 184, 10.1038/s41592-019-0666-6 · DOI
  20. 10.15252/MSB.202211081 · DOI · Paper page
  21. 2023, Alphafold2 versus experimental structures: evaluation on g protein-coupled receptors, Acta Pharmacol. Sinica, 44, 1, 10.1038/s41401-022-00938-y · DOI
  22. 2021, Graphdta: predicting drug–target binding affinity with graph neural networks, Bioinformatics, 37, 1140, 10.1093/bioinformatics/btaa921 · DOI
  23. Corso, G. , Stärk, H. , Jing, B. , Barzilay, R. & Jaakkola, T. S. Diffdock: Diffusion steps, twists, and turns for molecular docking. In The Eleventh International Conference on Learning Representations (2023).
  24. 2021, Multi-scale representation learning on proteins, Adv. Neural Inf. Process. Syst, 34, 25244
  25. 2020, Principal neighbourhood aggregation for graph nets, Adv. Neural Inf. Process. Syst, 33, 13260
  26. 10.1093/bioinformatics/btu626 · DOI
  27. 2018, Comparative assessment of scoring functions: the casf-2016 update, J. Chem. Inf. Model, 59, 895
  28. Stärk, H. , Ganea, O. , Pattanaik, L. , Barzilay, R. & Jaakkola, T. Equibind: Geometric deep learning for drug binding structure prediction. In International Conference on Machine Learning, 20503–20521 (PMLR, 2022).
  29. 10.1093/bioinformatics/bty374 · DOI
  30. 2019, Onionnet: a multiple-layer intermolecular-contact-based convolutional neural network for protein–ligand binding affinity prediction, ACS Omega, 4, 15956, 10.1021/acsomega.9b01997 · DOI
  31. 2021, Interactiongraphnet: A novel and efficient deep graph representation learning framework for accurate protein–ligand interaction predictions, J. Medicinal Chem, 64, 18209, 10.1021/acs.jmedchem.1c01830 · DOI
  32. 10.1021/ci300604z · DOI
  33. 2021, Gnina 1.0: molecular docking with deep learning, J. Cheminformatics, 13, 1
  34. Sverrisson, F. , Feydy, J. , Correia, B. E. & Bronstein, M. M. Fast end-to-end learning on protein surfaces. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 15272–15281 (2021). · DOI
  35. 10.1093/bioinformatics/btaa880 · DOI
  36. 10.1093/nar/gkl999 · DOI
  37. 10.1093/bioinformatics/btv256 · DOI
  38. 10.1093/nar/gkac1077 · DOI
  39. 2017, Excape-db: an integrated large scale dataset facilitating big data analysis in chemogenomics, J. Cheminformatics, 9, 1
  40. 2023, Papyrus: a large-scale curated dataset aimed at bioactivity predictions, J. Cheminformatics, 15, 3, 10.1186/s13321-022-00672-x · DOI
  41. 10.1093/NAR/GKAB294 · DOI
  42. 2019, Discovery of potent and selective covalent protein arginine methyltransferase 5 (prmt5) inhibitors, ACS Medicinal Chem. Lett, 10, 1033, 10.1021/acsmedchemlett.9b00074 · DOI
  43. 2019, Hiv-1 protease inhibitors incorporating stereochemically defined p2 ligands to optimize hydrogen bonding in the substrate envelope, J. Medicinal Chem, 62, 8062, 10.1021/acs.jmedchem.9b00838 · DOI
  44. 10.1016/j.tim.2016.03.010 · DOI
  45. 10.1038/s41586-018-0236-6 · DOI
  46. 10.1124/mol.116.105007 · DOI
  47. 10.1021/jm030331x · DOI
  48. 10.1039/C7MD00381A · DOI
  49. 2018, P2rank: machine learning based tool for rapid and accurate prediction of ligand binding sites from protein structure, J. Cheminformatics, 10, 1
  50. Bianchi, F. M. , Grattarola, D. & Alippi, C. Spectral clustering with graph neural networks for graph pooling. In International Conference on Machine Learning, 874–883 (PMLR, 2020).
  51. 1998, Feature trees: a new molecular similarity measure based on tree matching, J. Comput. Mol. Des, 12, 471, 10.1023/A:1008068904628 · DOI
  52. Jin, W. , Yang, K. , Barzilay, R. & Jaakkola, T. Learning multimodal graph-to-graph translation for molecule optimization. In International Conference on Learning Representations (2019).
  53. Cai, T. et al. Graphnorm: A principled approach to accelerating graph neural network training. In International Conference on Machine Learning, 1204–1215 (PMLR, 2021).
  54. Kingma, D. P. & Ba, J. Adam: A method for stochastic optimization. In International Conference on Learning Representations (2015).
  55. Loshchilov, I. & Hutter, F. Decoupled weight decay regularization. In International Conference on Learning Representations (2019).
  56. 10.1038/nbt.3988 · DOI · Paper page
  57. 10.1371/journal.pcbi.1003321 · DOI
  58. 10.1007/BF00342633 · DOI
  59. Vaswani, A. et al. Attention is all you need. Adv. Neural Inf. Process. Syst. 30 (2017).
  60. 10.1016/j.bcp.2015.11.013 · DOI
  61. Ismail-Fawaz, A. et al. An approach to multiple comparison benchmark evaluations that is stable under manipulation of the comparate set. arXiv preprint arXiv:2305.11921 (2023). · arXiv
  62. 2020, An open source chemical structure curation pipeline using rdkit, J. Cheminformatics, 12, 1
  63. 2019, Structure and energetics of ligand–fluorine interactions with galectin-3 backbone and side-chain amides: Insight into solvation effects and multipolar interactions, ChemMedChem, 14, 1528, 10.1002/cmdc.201900293 · DOI
  64. 2019, Substituted polyfluoroaryl interactions with an arginine side chain in galectin-3 are governed by steric-, desolvation and electronic conjugation effects, Org. & Biomol. Chem, 17, 1081, 10.1039/C8OB02888E · DOI
  65. Rao, R. et al. Evaluating protein transfer learning with tape. Adv. Neural Inf. Process. Syst. 32 (2019). · DOI · Paper page
  66. 2013, Some case studies on application of “rm2” metrics for judging quality of quantitative structure–activity relationship predictions: emphasis on scaling of response data, J. Comput. Chem, 34, 1071, 10.1002/jcc.23231 · DOI

Metadata

Authors
Koh, H. Y., Nguyen, A. T. N., Pan, S., May, L. T., & Webb, G. I.
Affiliations
  • Monash University (AU)
  • Griffith University (AU)
Corresponding author
  • Nguyen, Anh T.N.
  • Pan, Shirui
  • May, Lauren T.
  • Webb, Geoffrey I.
ORCID
Venue
Nature Machine Intelligence
Publisher
openRxiv
Published
2023-09-19
Volume
6
Issue
6
Pages
673-687
ISSN
2522-5839 (ISSN-L)
DOI
10.1101/2023.09.17.558145
OpenAlex
W4399746375
Cited by
108
References
1
Citations by year
2026: 46 · 2025: 48 · 2024: 13 · 2023: 1
FWCI
22.68
Type
journal article
Language
en
Open access
Licence
cc-by-nc-nd
Locations
  • Nature Machine Intelligence (publishedVersion)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
  • Griffith Research Online (submittedVersion, other-oa)
Funders
  • Australian Government
  • Monash University
  • Australian Research Data Commons
  • Medical Research Council
  • National Health and Medical Research Council (2013629)
  • Medical Research Future Fund
Topics
  • Life Sciences › Pharmacology, Toxicology and Pharmaceutics › Drug Discovery › Computational Drug Discovery Methods
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Interaction Studies and Fluorescence Analysis
Keywords
protein-ligand interactions, graph neural networks, sequence-based prediction, binding affinity prediction, interaction fingerprints, adenosine A1 receptor agonist, virtual screening, interpretability, ligand selectivity, drug discovery

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