Student or educator? Sign up with your school email and get 6 months of Pro free.
PreprintCited by 25

The DOI matches its Crossref record; 8 of the 8 references it cites were found in a registry or an index.

Simple computational methods can outperform deep learning in designing diverse, binder-enriched antibody libraries

Chinery, Lewis, Hummer, Alissa M., Mehta, Brij Bhushan, Akbar, Rahmad, Rawat, Puneet, Slabodkin, Andrei, Le Quy, Khang, Lund-Johansen, Fridtjof, Greiff, Victor, Jeliazkov, Jeliazko R., Deane, Charlotte M.·2024

Abstract

Strong antibody-antigen binding is the primary consideration when developing an efficacious therapeutic antibody. In recent years, much work has been devoted to applying complex machine learning models to this cause, yet simple baselines are often lacking. Here, we show that the widely used sequence alignment method, BLOSUM, can yield diverse, binder-enriched libraries from a single starting antibody. Using Trastuzumab-HER2 as a model system, we experimentally validated 720 novel designs generated with five different computational methods using surface plasmon resonance. The BLOSUM substitution matrix outperformed all four deep learning design approaches tested, achieving an estimated minimum binder enrichment of 12.5% and producing nine sub-nanomolar binders. These results underscore the importance of comparing against simple baselines and set a benchmark to guide future computational antibody library design. Abstract Figure:

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 · 10.1101/2024.03.26.586756

Cites

8 of 8 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

Show all 8 references

Research works printed but not linked: 26

  1. 10.1038/S41467-023-41205-W;TECHMETA · doi:10.1038/s41467-023-41205-w;techmeta
  2. Frey, N. C. et al. Protein discovery with discrete walk-jump sampling (2023). URL https://arxiv.org/abs/2306.12360v2. · arXiv
  3. 10.1073/pnas.2016239118 · DOI · Paper page
  4. 2021, Language models enable zero-shot prediction of the effects of mutations on protein function, Advances in Neural Information Processing Systems, 34, 29287
  5. Hie, B. L. et al. Efficient evolution of human antibodies from general protein language models. Nature Biotechnology 2023 1–9 (2023). URL https://www.nature.com/articles/s41587-023-01763-2.
  6. Shanehsazzadeh, A. et al. Unlocking de novo antibody design with generative artificial intelligence. bioRxiv 2023.01.08.523187 (2023). URL https://www.biorxiv.org/content/10. 1101/2023.01.08.523187v3. · doi:10.58599/ijsmien.2023.1801
  7. 10.1126/science.ade2574 · doi:10.1126/science.ade2574 · Paper page
  8. Olsen, T. H. , Moal, I. H. & Deane, C. M. Ablang: an antibody language model for completing antibody sequences. Bioinformatics Advances 2 (2022). URL https://academic.oup.com/bioinformaticsadvances/article/2/1/vbac046/6609807. · DOI · Paper page
  9. Olsen, T. H. , Moal, I. H. & Deane, C. M. Addressing the antibody germline bias and its effect on language models for improved antibody design. bioRxiv 2024.02.02.578678 (2024). URL https://www.biorxiv.org/content/10.1101/2024.02.02.578678v1 https://www.biorxiv.org/content/10.1101/2024.02.02.578678v1.abstract. · doi:10.1101/2024.02.02.578678v1
  10. 10.1093/bioadv/vbae202 · DOI
  11. Dreyer, F. A. , Cutting, D. , Schneider, C. , Kenlay, H. & Deane, C. M. Inverse folding for antibody sequence design using deep learning. ICML Workshop on Computational Biology (2023).
  12. Sun, Z.-Y. et al. Antibmpnn: Structure-guided graph neural networks for precision antibody engineering. Advanced Science n/a, e04278. URL https://advanced.onlinelibrary.wiley.com/doi/abs/10.1002/advs.202504278. · DOI
  13. 10.1002/pro.2829 · DOI · Paper page
  14. 10.1093/bioinformatics/bty635 · DOI · Paper page
  15. Chungyoun, M. , Ruffolo, J. & Gray, J. J. Flab: Benchmarking tasks in fitness landscape inference for antibodies. bioRxiv (2023). URL https://www.biorxiv.org/content/10.1101/2024.01.13.575504v1. · doi:10.1101/2024.01.13.575504v1
  16. 10.1073/pnas.89.22.10915 · DOI · Paper page
  17. Abanades, B. et al. Immunebuilder: Deep-learning models for predicting the structures of immune proteins. bioRxiv 2022.11.04.514231 (2022). URL https://www.biorxiv.org/content/10.1101/2022.11.04.514231v1 https://www.biorxiv.org/content/10.1101/2022.11.04.514231v1.abstract. · doi:10.1101/2022.11.04.514231v1
  18. 2001, Clinical trials of herceptin® (trastuzumab), European Journal of Cancer, 37, 18, 10.1016/S0959-8049(00)00404-4 · DOI · Paper page
  19. 10.1038/s41598-019-52507-9 · DOI
  20. Uçar, T. & Sormanni, P. Blosum is all you learn — generative antibody models reflect evolutionary priors. bioRxiv (2025). URL https://www.biorxiv.org/content/early/2025/10/27/2025.10.26.684652.1.
  21. 10.1371/journal.pone.0049535 · doi:10.1371/journal.pone.0049535
  22. 2006, Isolating and engineering human antibodies using yeast surface display, Nature Protocols 2006 1:2, 1, 755
  23. 10.1038/nbt0804-1035 · DOI · Paper page
  24. 10.1093/nar/gkab1050 · DOI
  25. 10.1002/pro.4205 · DOI · Paper page
  26. 10.1080/19420862.2021.2002236 · DOI

Metadata

Authors
Chinery, L., Hummer, A. M., Mehta, B. B., Akbar, R., Rawat, P., Slabodkin, A., Le Quy, K., Lund-Johansen, F., Greiff, V., Jeliazkov, J. R., & Deane, C. M.
Publisher
openRxiv
Published
2024-03-29
Cited by
25
References
8
Type
Preprint
Funders
  • Biotechnology and Biological Sciences Research Council (BB/V509681/1)
  • Leona M. and Harry B. Helmsley Charitable Trust (2019PG-T1D011)
  • UiO World-Leading Research Community
  • UiO: LifeScience Convergence Environment Immunolingo
  • European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement
  • EU Horizon 2020 iReceptorplus (825821)
  • Norwegian Cancer Society (215817)
  • The Research Council of Norway (300740, 331890)
  • Research Council of Norway IKTPLUSS project (311341)
  • Innovative Medicines Initiative 2 Joint Undertaking (101007799)
  • European Union (101125630)

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

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Simple computational methods can outperform deep learning in designing diverse, binder-enriched antibody libraries. https://app.cochat.ai/papers/10.1101/2024.03.26.586756
The paper
Chinery, L., Hummer, A. M., Mehta, B. B., Akbar, R., Rawat, P., Slabodkin, A., Le Quy, K., Lund-Johansen, F., Greiff, V., Jeliazkov, J. R., & Deane, C. M. (2024). Simple computational methods can outperform deep learning in designing diverse, binder-enriched antibody libraries. https://doi.org/10.1101/2024.03.26.586756

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.