Evaluation of AlphaFold antibody–antigen modeling with implications for improving predictive accuracy
Open accessPublished version of a preprint
Abstract
High resolution antibody-antigen structures provide critical insights into immune recognition and can inform therapeutic design. The challenges of experimental structural determination and the diversity of the immune repertoire underscore the necessity of accurate computational tools for modeling antibody-antigen complexes. Initial benchmarking showed that despite overall success in modeling protein-protein complexes, AlphaFold and AlphaFold-Multimer have limited success in modeling antibody-antigen interactions. In this study, we performed a thorough analysis of AlphaFold's antibody-antigen modeling performance on 427 nonredundant antibody-antigen complex structures, identifying useful confidence metrics for predicting model quality, and features of complexes associated with improved modeling success. Notably, we found that the latest version of AlphaFold improves near-native modeling success to over 30%, versus approximately 20% for a previous version, while increased AlphaFold sampling gives approximately 50% success. With this improved success, AlphaFold can generate accurate antibody-antigen models in many cases, while additional training or other optimization may further improve performance.
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?
Yin et al benchmarked 152 heterodimeric complexes and found AF2 and AF2-Multimer had a 51% success rate.
“Yin et al benchmarked 152 heterodimeric protein complexes revealing that AF2 and AF2-Multimer had a 51% success rate54,55”
From The power and pitfalls of AlphaFold2 for structure prediction beyond rigid globular proteins (2024) · cites this paper as [55]
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
ColabFold - Making protein folding accessible to all
Mirdita, M., Schütze, K., Moriwaki, Y., Heo, L., Ovchinnikov, S., & Steinegger, M. (2021) · doi:10.1101/2021.08.15.456425 · Cited by 9,391 · see citing works
Protein complex prediction with AlphaFold-Multimer
Evans, R., O’Neill, M., Pritzel, A., Antropova, N., Senior, A., Green, T., Žídek, A., Bates, R., Blackwell, S., Yim, J., et al. (2021) · doi:10.1101/2021.10.04.463034 · Cited by 3,808 · see citing works
SARS-CoV-2 neutralizing antibody structures inform therapeutic strategies
Barnes, C. O., Jette, C. A., Abernathy, M. E., Dam, K.-M. A., Esswein, S. R., Gristick, H. B., Malyutin, A. G., Sharaf, N. G., Huey-Tubman, K. E., Lee, Y. E., et al. (2020) · Nature · doi:10.1038/s41586-020-2852-1 · Cited by 1,599 · see citing works
Improved prediction of protein-protein interactions using AlphaFold2
Bryant, P., Pozzati, G., & Elofsson, A. (2021) · doi:10.1101/2021.09.15.460468 · Cited by 1,025 · see citing works
The promise and challenge of high-throughput sequencing of the antibody repertoire
Georgiou, G., Ippolito, G. C., Beausang, J., Busse, C. E., Wardemann, H., & Quake, S. R. (2014) · Nature Biotechnology · doi:10.1038/nbt.2782 · Cited by 652 · see citing works
Ahdritz, G., Bouatta, N., Floristean, C., Kadyan, S., Xia, Q., Gerecke, W., O’Donnell, T. J., Berenberg, D., Fisk, I., Zanichelli, N., et al. (2024) · Nature Methods · doi:10.1038/s41592-024-02272-z · Cited by 439 · see citing works
Vreven, T., Moal, I. H., Vangone, A., Pierce, B. G., Kastritis, P. L., Torchala, M., Chaleil, R., Jiménez-García, B., Bates, P. A., Fernandez-Recio, J., et al. (2015) · Journal of Molecular Biology · doi:10.1016/j.jmb.2015.07.016 · Cited by 424 · see citing works
Self-citationBenchmarking AlphaFold for protein complex modeling reveals accuracy determinants
Yin, R., Feng, B. Y., Varshney, A., & Pierce, B. G. (2021) · doi:10.1101/2021.10.23.465575 · Cited by 335 · see citing works
Self-citationStructure-Based Vaccine Antigen Design
Graham, B. S., Gilman, M. S. A., & McLellan, J. S. (2019) · Annual Review of Medicine · doi:10.1146/annurev-med-121217-094234 · Cited by 234 · see citing works
Krawczyk, K., Baker, T., Shi, J., & Deane, C. M. (2013) · Protein Engineering Design and Selection · doi:10.1093/protein/gzt043 · Cited by 87 · see citing works
Improved Multimer Prediction using Massive Sampling with AlphaFold in CASP15
Wallner, B. (2023) · doi:10.22541/au.168170986.60158084/v1 · Cited by 53 · see citing works
Flagged first, then by citations
Research works printed but not linked: 44
- Abanades B, Wong WK, Boyles F, Georges G, Bujotzek A, Deane CM. ImmuneBuilder: deep‐learning models for predicting the structures of immune proteins. Commun Biol. 2023; 6(1): 575. 37248282 PMC10227038
- Alford RF, Leaver‐Fay A, Jeliazkov JR, O'Meara MJ, DiMaio FP, Park H, et al. The Rosetta all‐atom energy function for macromolecular modeling and design. J Chem Theory Comput. 2017; 13(6): 3031– 3048. 28430426 PMC5717763
- Ambrosetti F, Jimenez‐Garcia B, Roel‐Touris J, Bonvin A. Modeling antibody‐antigen complexes by information‐driven docking. Structure. 2020; 28(1): 119– 129.e2. 31727476
- Bo C, Ziwei X, Jiezhong Q, Zhaofeng Y, Jinbo X, Jie T. Improve the protein complex prediction with protein language models. bioRxiv. 2022. 10.1101/2022.09.15.508065 · DOI
- Brenke R, Hall DR, Chuang GY, Comeau SR, Bohnuud T, Beglov D, et al. Application of asymmetric statistical potentials to antibody‐protein docking. Bioinformatics. 2012; 28(20): 2608– 2614. 23053206 PMC3467743
- Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos J, Bealer K, et al. BLAST+: architecture and applications. BMC Bioinformatics. 2009; 10: 421. 20003500 PMC2803857
- Carter PJ. Potent antibody therapeutics by design. Nat Rev Immunol. 2006; 6(5): 343– 357. 16622479
- Chothia C, Lesk AM. Canonical structures for the hypervariable regions of immunoglobulins. J Mol Biol. 1987; 196(4): 901– 917. 3681981
- Dreyfus C, Laursen NS, Kwaks T, Zuijdgeest D, Khayat R, Ekiert DC, et al. Highly conserved protective epitopes on influenza B viruses. Science. 2012; 337(6100): 1343– 1348. 22878502 PMC3538841
- Dunbar J, Deane CM. ANARCI: antigen receptor numbering and receptor classification. Bioinformatics. 2016; 32(2): 298– 300. 26424857 PMC4708101
- Dunbar J, Krawczyk K, Leem J, Baker T, Fuchs A, Georges G, et al. SAbDab: the structural antibody database. Nucleic Acids Res. 2014; 42(Database issue): D1140– D1146. 24214988 PMC3965125
- Fu L, Niu B, Zhu Z, Wu S, Li W. CD‐HIT: accelerated for clustering the next‐generation sequencing data. Bioinformatics. 2012; 28(23): 3150– 3152. 23060610 PMC3516142
- Guest JD, Vreven T, Zhou J, Moal I, Jeliazkov JR, Gray JJ, et al. An expanded benchmark for antibody‐antigen docking and affinity prediction reveals insights into antibody recognition determinants. Structure. 2021; 29(6): 606– 621.e5. 33539768 PMC8178184
- Haidar JN, Yuan QA, Zeng L, Snavely M, Luna X, Zhang H, et al. A universal combinatorial design of antibody framework to graft distinct CDR sequences: a bioinformatics approach. Proteins. 2012; 80(3): 896– 912. 22180101
- Hanf KJ, Arndt JW, Chen LL, Jarpe M, Boriack‐Sjodin PA, Li Y, et al. Antibody humanization by redesign of complementarity‐determining region residues proximate to the acceptor framework. Methods. 2014; 65(1): 68– 76. 23816785
- Hie BL, Shanker VR, Xu D, Bruun TUJ, Weidenbacher PA, Tang S, et al. Efficient evolution of human antibodies from general protein language models. Nat Biotechnol. 2023. 10.1038/s41587-023-01763-2 PMC10869273 37095349 · DOI
- Honegger A, Pluckthun A. Yet another numbering scheme for immunoglobulin variable domains: an automatic modeling and analysis tool. J Mol Biol. 2001; 309(3): 657– 670. 11397087
- Johansson‐Akhe I, Wallner B. Improving peptide‐protein docking with AlphaFold‐multimer using forced sampling. Front Bioinform. 2022; 2: 959160. 36304330 PMC9580857
- Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, et al. Applying and improving AlphaFold at CASP14. Proteins. 2021b; 89: 1711– 1721. 34599769 PMC9299164
- Kappler K, Hennet T. Emergence and significance of carbohydrate‐specific antibodies. Genes Immun. 2020; 21(4): 224– 239. 32753697 PMC7449879
- Lee JH, Yin R, Ofek G, Pierce BG. Structural features of antibody‐peptide recognition. Front Immunol. 2022; 13: 910367. 35874680 PMC9302003
- Leman JK, Weitzner BD, Lewis SM, Adolf‐Bryfogle J, Alam N, Alford RF, et al. Macromolecular modeling and design in Rosetta: recent methods and frameworks. Nat Methods. 2020; 17(7): 665– 680. 32483333 PMC7603796
- Lensink M, Brysbaert G, Raouraoua N, Bates PA, Giulini M, Honorato RV, et al. Impact of AlphaFold on structure prediction of protein complexes: the CASP15‐CAPRI experiment. Proteins. 2023; 91(12): 1658– 1683. 37905971 PMC10841881
- Lensink MF, Nadzirin N, Velankar S, Wodak SJ. Modeling protein‐protein, protein‐peptide, and protein‐oligosaccharide complexes: CAPRI 7th edition. Proteins. 2020; 88(8): 916– 938. 31886916
- Li Y, Li H, Yang F, Smith‐Gill SJ, Mariuzza RA. X‐ray snapshots of the maturation of an antibody response to a protein antigen. Nat Struct Biol. 2003; 10(6): 482– 488. 12740607
- Li Z, Woo CJ, Iglesias‐Ussel MD, Ronai D, Scharff MD. The generation of antibody diversity through somatic hypermutation and class switch recombination. Genes Dev. 2004; 18(1): 1– 11. 14724175
- Motmaen A, Dauparas J, Baek M, Abedi MH, Baker D, Bradley P. Peptide‐binding specificity prediction using fine‐tuned protein structure prediction networks. Proc Natl Acad Sci U S A. 2023; 120(9): e2216697120. 36802421 PMC9992841
- Nelson AL, Dhimolea E, Reichert JM. Development trends for human monoclonal antibody therapeutics. Nat Rev Drug Discov. 2010; 9(10): 767– 774. 20811384
- Olsen TH, Moal IH, Deane CM. AbLang: an antibody language model for completing antibody sequences. Bioinform Adv. 2022; 2(1): vbac046. 36699403 PMC9710568
- Pierce BG, Hourai Y, Weng Z. Accelerating protein docking in ZDOCK using an advanced 3D convolution library. PloS One. 2011; 6(9): e24657. 21949741 PMC3176283
- Rappuoli R, Bottomley MJ, D'Oro U, Finco O, De Gregorio E. Reverse vaccinology 2.0: human immunology instructs vaccine antigen design. J Exp Med. 2016; 213(4): 469– 481. 27022144 PMC4821650
- Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez JC, et al. pROC: an open‐source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics. 2011; 12: 77. 21414208 PMC3068975
- Ruffolo JA, Chu LS, Mahajan SP, Gray JJ. Fast, accurate antibody structure prediction from deep learning on massive set of natural antibodies. Nat Commun. 2023; 14(1): 2389. 37185622 PMC10129313
- Ruffolo JA, Gray JJ, Sulam J. Deciphering antibody affinity maturation with language models and weakly supervised learning. arXiv preprint. arXiv:211207782. 2021.
- Scott AM, Wolchok JD, Old LJ. Antibody therapy of cancer. Nat Rev Cancer. 2012; 12(4): 278– 287. 22437872
- Sela‐Culang I, Kunik V, Ofran Y. The structural basis of antibody‐antigen recognition. Front Immunol. 2013; 4: 302. 24115948 PMC3792396
- Sircar A, Gray JJ. SnugDock: paratope structural optimization during antibody‐antigen docking compensates for errors in antibody homology models. PLoS Comput Biol. 2010; 6(1): e1000644. 20098500 PMC2800046
- Stranges PB, Kuhlman B. A comparison of successful and failed protein interface designs highlights the challenges of designing buried hydrogen bonds. Protein Sci. 2013; 22(1): 74– 82. 23139141 PMC3575862
- Wallner B. AFsample: improving multimer prediction with AlphaFold using massive sampling. Bioinformatics. 2023a; 39(9): btad573. 37713472 PMC10534052
- Wickham H. ggplot2: elegant graphics for data analysis. New York: Springer‐Verlag; 2016.
- Zheng W, Li Y, Zhang C, Zhou X, Pearce R, Bell EW, et al. Protein structure prediction using deep learning distance and hydrogen‐bonding restraints in CASP14. Proteins. 2021; 89(12): 1734– 1751. 34331351 PMC8616857
- Zhou T, Lynch RM, Chen L, Acharya P, Wu X, Doria‐Rose NA, et al. Structural repertoire of HIV‐1‐neutralizing antibodies targeting the CD4 supersite in 14 donors. Cell. 2015; 161(6): 1280– 1292. 26004070 PMC4683157
- Zhu J, Weng Z. FAST: a novel protein structure alignment algorithm. Proteins. 2005; 58(3): 618– 627. 15609341
- Ziyao L, Xuyang L, Weijie C, Shen F, Bi H, Ke G, et al. Uni‐fold: an open‐source platform for developing protein folding models beyond AlphaFold. bioRxiv. 2022. 10.1101/2022.08.04.502811 · DOI
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
“Yin et al. 55,54 benchmarked 152 heterodimeric protein complexes, revealing that AF2 and AlphaFold2-Multimer had a 51% success rate.”
BackgroundThe Virtual Lab of AI agents designs new SARS-CoV-2 nanobodies
Swanson, K., Wu, W., Bulaong, N. L., Pak, J. E., & Zou, J. (2025) · Nature · doi:10.1038/s41586-025-09442-9 · Cited by 169 · see citing works
The Virtual Lab: AI Agents Design New SARS-CoV-2 Nanobodies with Experimental Validation
Swanson, K., Wu, W., Bulaong, N. L., Pak, J. E., & Zou, J. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.11.11.623004 · Cited by 56 · see citing works
Wu, D., Yin, R., Chen, G., Ribeiro-Filho, H. V., Cheung, M., Robbins, P. F., Mariuzza, R. A., & Pierce, B. G. (2024) · Science Advances · doi:10.1126/sciadv.adq6150 · Cited by 27 · see citing works
Self-citationLiu, L., Liu, Y., Ye, X., Zhang, S., Li, Y., Zhu, K., Xue, Y., Zhou, J., Zhang, X., & Fang, X. (2026) · Nature Machine Intelligence · doi:10.1038/s42256-026-01264-2 · Cited by 1 · see citing works
- Authors
- Yin, R., & Pierce, B. G.
- Affiliations
- Institute for Bioscience and Biotechnology Research (US)
- Corresponding author
- Pierce, Brian G.
- Venue
- Protein Science
- Publisher
- Wiley
- Published
- 2023-12-27
- Volume
- 33
- Issue
- 1
- Pages
- e4865-e4865
- ISSN
- 0961-8368 (ISSN-L)
- 1469-896X
- PMC
- PMC10751731
- OpenAlex
- W4389568864
- Cited by
- 145
- References
- 12
- Citations by year
- 2026: 47 · 2025: 58 · 2024: 37 · 2023: 2 · 2022: 1
- FWCI
- 12.81
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Protein Science (publishedVersion, cc-by)
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- University Libraries (University of Maryland) (submittedVersion)
- PubMed Central (submittedVersion, cc-by)
- Digital Repository at the University of Maryland (University of Maryland College Park)
- Funders
- National Institutes of Health (R35 GM144083)
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › vaccines and immunoinformatics approaches
- Physical Sciences › Materials Science › Materials Chemistry › Enzyme Structure and Function
- Keywords
- AlphaFold, AlphaFold-Multimer, immune complexes, immune recognition, benchmarking
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
