Peptide-binding specificity prediction using fine-tuned protein structure prediction networks
Open accessPublished version of a preprint
Abstract
Peptide-binding proteins play key roles in biology, and predicting their binding specificity is a long-standing challenge. While considerable protein structural information is available, the most successful current methods use sequence information alone, in part because it has been a challenge to model the subtle structural changes accompanying sequence substitutions. Protein structure prediction networks such as AlphaFold model sequence-structure relationships very accurately, and we reasoned that if it were possible to specifically train such networks on binding data, more generalizable models could be created. We show that placing a classifier on top of the AlphaFold network and fine-tuning the combined network parameters for both classification and structure prediction accuracy leads to a model with strong generalizable performance on a wide range of Class I and Class II peptide-MHC interactions that approaches the overall performance of the state-of-the-art NetMHCpan sequence-based method. The peptide-MHC optimized model shows excellent performance in distinguishing binding and non-binding peptides to SH3 and PDZ domains. This ability to generalize well beyond the training set far exceeds that of sequence-only models and should be particularly powerful for systems where less experimental data are available.
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.
Integrity signals
Registry
Cites
Statistics
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.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Several groups reported AF2 models with low PAE (<5 Å) that still deviated from experimental data.
“Several groups have reported cases of AF2 models with low PAE values (< 5 Å) that deviated from experimental data62,63”
From The power and pitfalls of AlphaFold2 for structure prediction beyond rigid globular proteins (2024) · cites this paper as [63]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Highly accurate protein structure prediction with AlphaFold
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., et al. (2021) · Nature · doi:10.1038/s41586-021-03819-2 · Cited by 43,764 · see citing works
Berman, H. M. (2000) · Nucleic Acids Research · doi:10.1093/nar/28.1.235 · Cited by 40,196 · see citing works
Biopython: freely available python tools for computational molecular biology and bioinformatics
Cock, P. J. A., Antao, T., Chang, J. T., Chapman, B. A., Cox, C. J., Dalke, A., Friedberg, I., Hamelryck, T., Kauff, F., Wilczynski, B., et al. (2009) · Bioinformatics · doi:10.1093/bioinformatics/btp163 · Cited by 6,080 · see citing works
Accurate prediction of protein structures and interactions using a 3-track network
Baek, M., DiMaio, F., Anishchenko, I., Dauparas, J., Ovchinnikov, S., Lee, G. R., Wang, J., Cong, Q., Kinch, L. N., Schaeffer, R. D., et al. (2021) · doi:10.1101/2021.06.14.448402 · Cited by 5,542 · see citing works
Self-citationReynisson, B., Alvarez, B., Paul, S., Peters, B., & Nielsen, M. (2020) · Nucleic Acids Research · doi:10.1093/nar/gkaa379 · Cited by 2,361 · see citing works
Abelin, J. G., Keskin, D. B., Sarkizova, S., Hartigan, C. R., Zhang, W., Sidney, J., Stevens, J. D., Lane, W. J., Zhang, G., Eisenhaure, T., et al. (2017) · Immunity · doi:10.1016/j.immuni.2017.02.007 · Cited by 593 · see citing works
Computed structures of core eukaryotic protein complexes
Humphreys, I. R., Pei, J., Baek, M., Krishnakumar, A., Anishchenko, I., Ovchinnikov, S., Zhang, J., Ness, T. J., Banjade, S., Bagde, S. R., et al. (2021) · Science · doi:10.1126/science.abm4805 · Cited by 542 · see citing works
Self-citationSarkizova, S., Klaeger, S., Le, P. M., Li, L. W., Oliveira, G., Keshishian, H., Hartigan, C. R., Zhang, W., Braun, D. A., Ligon, K. L., et al. (2019) · Nature Biotechnology · doi:10.1038/s41587-019-0322-9 · Cited by 505 · see citing works
A Specificity Map for the PDZ Domain Family
Tonikian, R., Zhang, Y., Sazinsky, S. L., Currell, B., Yeh, J.-H., Reva, B., Held, H. A., Appleton, B. A., Evangelista, M., Wu, Y., et al. (2008) · PLoS Biology · doi:10.1371/journal.pbio.0060239 · Cited by 476 · see citing works
Reynisson, B., Barra, C., Kaabinejadian, S., Hildebrand, W. H., Peters, B., & Nielsen, M. (2020) · Journal of Proteome Research · doi:10.1021/acs.jproteome.9b00874 · Cited by 448 · see citing works
Improving de novo Protein Binder Design with Deep Learning
(2022) · doi:10.1101/2022.06.15.495993 · Cited by 402 · see citing works
Applying and improving AlphaFold at CASP14
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., et al. (2021) · Proteins: Structure, Function, and Bioinformatics · doi:10.1002/prot.26257 · Cited by 370 · see citing works
Biophysical prediction of protein–peptide interactions and signaling networks using machine learning
Cunningham, J. M., Rogers, J. R., Koytiger, G., Sorger, P. K., & AlQuraishi, M. (2020) · Nature Methods · doi:10.1038/s41592-019-0687-1 · Cited by 108 · see citing works
Smith, C. A., & Kortemme, T. (2010) · Journal of Molecular Biology · doi:10.1016/j.jmb.2010.07.032 · Cited by 103 · see citing works
Venkatesh, G. S., Grover, A., Srinivasaraghavan, G., & Rao, S. (2020) · Bioinformatics · doi:10.1093/bioinformatics/btaa479 · Cited by 52 · see citing works
General Prediction of Peptide-MHC Binding Modes Using Incremental Docking: A Proof of Concept
Antunes, D. A., Devaurs, D., Moll, M., Lizee, G. A., & Kavraki, L. E. (2018) · Scientific Reports · doi:10.1038/s41598-018-22173-4 · Cited by 51 · see citing works
Khan, J. M., & Ranganathan, S. (2010) · Immunome Research · doi:10.1186/1745-7580-6-s1-s2 · Cited by 49 · see citing works
Jin, J., Liu, Z., Nasiri, A., Cui, Y., Louis, S., Zhang, A., Zhao, Y., & Hu, J. (2021) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.26065 · Cited by 41 · see citing works
DeepLigand: accurate prediction of MHC class I ligands using peptide embedding
Zeng, H., & Gifford, D. K. (2019) · Bioinformatics · doi:10.1093/bioinformatics/btz330 · Cited by 40 · see citing works
Subangstrom Accuracy in pHLA-I Modeling by Rosetta FlexPepDock Refinement Protocol
Liu, T., Pan, X., Chao, L., Tan, W., Qu, S., Yang, L., Wang, B., & Mei, H. (2014) · Journal of Chemical Information and Modeling · doi:10.1021/ci500393h · Cited by 34 · see citing works
Large‐scale survey and database of high affinity ligands for peptide recognition modules
Teyra, J., Kelil, A., Jain, S., et al. (2020) · Molecular Systems Biology · doi:10.15252/msb.20199310 · Cited by 33 · see citing works
AlphaFold encodes the principles to identify high affinity peptide bindersPreprint
Chang, L., & Pérez, A. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.03.18.484931 · Cited by 27 · see citing works
Structural Prediction of Peptide–MHC Binding Modes
Perez, M. A. S., Cuendet, M. A., Röhrig, U. F., Michielin, O. A., & Zoete, V. (2022) · Methods in molecular biology · doi:10.1007/978-1-0716-1855-4_13 · Cited by 26 · see citing works
Accurate structure prediction of peptide–MHC complexes for identifying highly immunogenic antigens
Park, M.-S., Park, S. Y., Miller, K. R., et al. (2013) · Molecular Immunology · doi:10.1016/j.molimm.2013.04.011 · Cited by 16 · see citing works
Predictions of immunogenicity reveal potent SARS-CoV-2 CD8+ T-cell epitopesPreprint
Gfeller, D., Schmidt, J., Croce, G., Guillaume, P., Bobisse, S., Genolet, R., Queiroz, L., Cesbron, J., Racle, J., & Harari, A. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.05.23.492800 · Cited by 14 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
The power and pitfalls of AlphaFold2 for structure prediction beyond rigid globular proteins
Agarwal, V., & McShan, A. C. (2024) · Nature Chemical Biology · doi:10.1038/s41589-024-01638-w · Cited by 155 · see citing works
“Several groups have reported cases of AF2 models with low PAE values ( 5 Å) that deviated from experimental data 62,63 .”
BackgroundFrom interaction networks to interfaces, scanning intrinsically disordered regions using AlphaFold2
Bret, H., Gao, J., Zea, D. J., Andreani, J., & Guerois, R. (2024) · Nature Communications · doi:10.1038/s41467-023-44288-7 · Cited by 103 · see citing works
Chang, L., Mondal, A., Singh, B., Martínez‐Noa, Y., & Perez, A. (2024) · WIREs Computational Molecular Science · doi:10.1002/wcms.1693 · Cited by 56 · see citing works
- Authors
- Motmaen, A., Dauparas, J., Baek, M., Abedi, M. H., Baker, D., & Bradley, P.
- Affiliations
- University of Washington (US)
- Howard Hughes Medical Institute (US)
- Corresponding author
- Baker, David
- Bradley, Philip
- Venue
- Proceedings of the National Academy of Sciences
- Publisher
- National Academy of Sciences
- Published
- 2023-02-21
- Volume
- 120
- Issue
- 9
- Pages
- e2216697120-e2216697120
- ISSN
- 0027-8424 (ISSN-L)
- 1091-6490
- PMC
- PMC9992841
- OpenAlex
- W4321440624
- Cited by
- 124
- References
- 25
- Citations by year
- 2026: 25 · 2025: 39 · 2024: 29 · 2023: 25 · 2022: 6
- FWCI
- 10.48
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Proceedings of the National Academy of Sciences (publishedVersion, cc-by)
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Funders
- Microsoft (NA)
- Howard Hughes Medical Institute (NA)
- HHS | National Institutes of Health (GM141457)
- Jane Coffin Childs Memorial Fund for Medical Research (NA)
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › vaccines and immunoinformatics approaches
- 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
- Keywords
- peptide-MHC interaction, AlphaFold, fine-tuning, binding affinity prediction, MHC class I, MHC class II, SH3 domain, PDZ domain, NetMHCpan
- MeSH terms
- Histocompatibility Antigens Class II, Genes, MHC Class II, Peptides, Protein Binding, PDZ Domains
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
