Atomically accurate de novo design of antibodies with RFdiffusion

Abstract

Despite the central role of antibodies in modern medicine, no method currently exists to design novel, epitope-specific antibodies entirely in silico. Instead, antibody discovery currently relies on immunization, random library screening or the isolation of antibodies directly from patients 1 . Here we demonstrate that combining computational protein design using a fine-tuned RFdiffusion 2 network with yeast display screening enables the de novo generation of antibody variable heavy chains (VHHs), single-chain variable fragments (scFvs) and full antibodies that bind to user-specified epitopes with atomic-level precision. We experimentally characterize VHH binders to four disease-relevant epitopes. Cryo-electron microscopy confirms the binding pose of designed VHHs targeting influenza haemagglutinin and Clostridium difficile toxin B (TcdB). A high-resolution structure of the influenza-targeting VHH confirms atomic accuracy of the designed complementarity-determining regions (CDRs). Although initial computational designs exhibit modest affinity (tens to hundreds of nanomolar K d ), affinity maturation using OrthoRep 3 enables production of single-digit nanomolar binders that maintain the intended epitope selectivity. We further demonstrate the de novo design of scFvs to TcdB and a PHOX2B peptide–MHC complex by combining designed heavy-chain and light-chain CDRs. Cryo-electron microscopy confirms the binding pose for two distinct TcdB scFvs, with high-resolution data for one design verifying the atomically accurate design of the conformations of all six CDR loops. Our approach establishes a framework for the computational design, screening and characterization of fully de novo antibodies with atomic-level precision in both structure and epitope targeting.

Shown for reading. Its publisher allows quoting it, not republishing it.

Full text

Its licence does not allow us to show the full text. Read it at PubMed Central:

References 58

1. 1.

   This reference’s wording can’t be shown under the paper’s licence.

   Not in our library10.1093/abt/tbac021

2. 2.

   This reference’s wording can’t be shown under the paper’s licence.

   Not in our library10.1038/nri3285

3. 3.

   This reference’s wording can’t be shown under the paper’s licence.

   Not in our library10.1073/pnas.1422401112

4. 4.

   This reference’s wording can’t be shown under the paper’s licence.

   Not in our library10.1038/srep41306

5. 5.

   This reference’s wording can’t be shown under the paper’s licence.

   Not in our library10.1016/b978-0-12-381270-4.00019-6

6. 6.

   This reference’s wording can’t be shown under the paper’s licence.

   Not in our library

7. 7.

   This reference’s wording can’t be shown under the paper’s licence.

   Not in our library10.1101/2022.12.22.521698

8. 8.

   This reference’s wording can’t be shown under the paper’s licence.

   Not in our library10.1101/2023.01.08.523187

9. 9.
   De novo design of protein structure and function with RFdiffusion · 2023

   No notice
10. 10.
    De novo design of high-affinity binders of bioactive helical peptides · 2023

    No notice
11. 11.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1101/2024.10.11.617496

12. 12.
    Design of protein-binding proteins from the target structure alone · 2022

    No notice
13. 13.
    De novo design of protein interactions with learned surface fingerprints · 2023

    No notice
14. 14.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1101/2024.09.30.615802

15. 15.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.48550/arxiv.2405.07622

16. 16.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1093/nar/28.1.235

17. 17.
    Highly accurate protein structure prediction with AlphaFold · 2021

    No notice
18. 18.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1073/pnas.1914677117

19. 19.
    Scaffolding protein functional sites using deep learning · 2022

    No notice
20. 20.
    Improving de novo protein binder design with deep learning · 2023

    No notice
21. 21.
    Evaluation of AlphaFold antibody–antigen modeling with implications for improving predictive accuracy · 2023

    No notice
22. 22.
    Accurate structure prediction of biomolecular interactions with AlphaFold 3 · 2024

    No notice
23. 23.
    NANOBODIES®: A Review of Diagnostic and Therapeutic Applications · 2023

    No notice
24. 24.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1186/s12976-015-0010-5

25. 25.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library

26. 26.
    Analysis of nanobody paratopes reveals greater diversity than classical antibodies · 2018

    No notice
27. 27.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1074/jbc.m806889200

28. 28.
    Robust deep learning–based protein sequence design using ProteinMPNN · 2022

    No notice
29. 29.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1126/scitranslmed.abn1252

30. 30.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1101/2024.08.26.609740

31. 31.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1126/science.adm9073

32. 32.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1016/j.cell.2018.10.021

33. 33.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1126/science.adf6591

34. 34.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1101/2023.10.18.562842

35. 35.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1016/j.yjsbx.2025.100120

36. 36.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1016/j.cell.2018.12.028

37. 37.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1038/s41586-023-06706-0

38. 38.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1126/sciimmunol.adj5792

39. 39.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1038/s41587-024-02505-8

40. 40.

    This reference’s wording can’t be shown under the paper’s licence.

    Not in our library10.1073/pnas.1921964117

Reads

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.