AlphaBind, a domain-specific model to predict and optimize antibody–antigen binding affinity
Antibodies are versatile therapeutic molecules that use combinatorial sequence diversity to cover a vast fitness landscape. Designing optimal antibody sequences, however, remains a major challenge. Recent advances in deep learning provide opportunities to address this challenge by learning sequence–...
Saved in:
| Main Authors: | , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2025-12-01
|
| Series: | mAbs |
| Subjects: | |
| Online Access: | https://www.tandfonline.com/doi/10.1080/19420862.2025.2534626 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849710222644346880 |
|---|---|
| author | Aditya A. Agarwal James Harrang David Noble Kerry L. McGowan Adrian W. Lange Emily Engelhart Miranda C. Lahman Jeffrey Adamo Xin Yu Oliver Serang Kyle J. Minch Kimberly Y. Wellman David A. Younger Randolph M. Lopez Ryan O. Emerson |
| author_facet | Aditya A. Agarwal James Harrang David Noble Kerry L. McGowan Adrian W. Lange Emily Engelhart Miranda C. Lahman Jeffrey Adamo Xin Yu Oliver Serang Kyle J. Minch Kimberly Y. Wellman David A. Younger Randolph M. Lopez Ryan O. Emerson |
| author_sort | Aditya A. Agarwal |
| collection | DOAJ |
| description | Antibodies are versatile therapeutic molecules that use combinatorial sequence diversity to cover a vast fitness landscape. Designing optimal antibody sequences, however, remains a major challenge. Recent advances in deep learning provide opportunities to address this challenge by learning sequence–function relationships to accurately predict fitness landscapes. These models enable efficient in silico prescreening and optimization of antibody candidates. By focusing experimental efforts on the most promising candidates guided by deep learning predictions, antibodies with optimal properties can be designed more quickly and effectively. Here we present AlphaBind, a domain-specific model that uses protein language model embeddings and pre-training on millions of quantitative laboratory measurements of antibody–antigen binding strength to achieve state-of-the-art performance for guided affinity optimization of parental antibodies. We demonstrate that an AlphaBind-powered antibody optimization pipeline can deliver candidates with substantially improved binding affinity across four parental antibodies (some of which were already affinity-matured) and using two different types of training data. The resulting candidates, which include up to 11 mutations from parental sequence, yield a sequence diversity that allows optimization of other biophysical characteristics, all while using only a single round of data generation for each parental antibody. AlphaBind weights and code are publicly available at: https://github.com/A-Alpha-Bio/alphabind. |
| format | Article |
| id | doaj-art-817a6ab3e13646e2b9ff5db4be02f10c |
| institution | DOAJ |
| issn | 1942-0862 1942-0870 |
| language | English |
| publishDate | 2025-12-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | mAbs |
| spelling | doaj-art-817a6ab3e13646e2b9ff5db4be02f10c2025-08-20T03:14:58ZengTaylor & Francis GroupmAbs1942-08621942-08702025-12-0117110.1080/19420862.2025.2534626AlphaBind, a domain-specific model to predict and optimize antibody–antigen binding affinityAditya A. Agarwal0James Harrang1David Noble2Kerry L. McGowan3Adrian W. Lange4Emily Engelhart5Miranda C. Lahman6Jeffrey Adamo7Xin Yu8Oliver Serang9Kyle J. Minch10Kimberly Y. Wellman11David A. Younger12Randolph M. Lopez13Ryan O. Emerson14Data Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAHealthcare & Life Sciences, Nvidia Corporation, Santa Clara, CA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAData Science, A-Alpha Bio Inc, Seattle, WA, USAAntibodies are versatile therapeutic molecules that use combinatorial sequence diversity to cover a vast fitness landscape. Designing optimal antibody sequences, however, remains a major challenge. Recent advances in deep learning provide opportunities to address this challenge by learning sequence–function relationships to accurately predict fitness landscapes. These models enable efficient in silico prescreening and optimization of antibody candidates. By focusing experimental efforts on the most promising candidates guided by deep learning predictions, antibodies with optimal properties can be designed more quickly and effectively. Here we present AlphaBind, a domain-specific model that uses protein language model embeddings and pre-training on millions of quantitative laboratory measurements of antibody–antigen binding strength to achieve state-of-the-art performance for guided affinity optimization of parental antibodies. We demonstrate that an AlphaBind-powered antibody optimization pipeline can deliver candidates with substantially improved binding affinity across four parental antibodies (some of which were already affinity-matured) and using two different types of training data. The resulting candidates, which include up to 11 mutations from parental sequence, yield a sequence diversity that allows optimization of other biophysical characteristics, all while using only a single round of data generation for each parental antibody. AlphaBind weights and code are publicly available at: https://github.com/A-Alpha-Bio/alphabind.https://www.tandfonline.com/doi/10.1080/19420862.2025.2534626Antibody engineeringcomputational protein designmachine learningyeast display |
| spellingShingle | Aditya A. Agarwal James Harrang David Noble Kerry L. McGowan Adrian W. Lange Emily Engelhart Miranda C. Lahman Jeffrey Adamo Xin Yu Oliver Serang Kyle J. Minch Kimberly Y. Wellman David A. Younger Randolph M. Lopez Ryan O. Emerson AlphaBind, a domain-specific model to predict and optimize antibody–antigen binding affinity mAbs Antibody engineering computational protein design machine learning yeast display |
| title | AlphaBind, a domain-specific model to predict and optimize antibody–antigen binding affinity |
| title_full | AlphaBind, a domain-specific model to predict and optimize antibody–antigen binding affinity |
| title_fullStr | AlphaBind, a domain-specific model to predict and optimize antibody–antigen binding affinity |
| title_full_unstemmed | AlphaBind, a domain-specific model to predict and optimize antibody–antigen binding affinity |
| title_short | AlphaBind, a domain-specific model to predict and optimize antibody–antigen binding affinity |
| title_sort | alphabind a domain specific model to predict and optimize antibody antigen binding affinity |
| topic | Antibody engineering computational protein design machine learning yeast display |
| url | https://www.tandfonline.com/doi/10.1080/19420862.2025.2534626 |
| work_keys_str_mv | AT adityaaagarwal alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT jamesharrang alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT davidnoble alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT kerrylmcgowan alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT adrianwlange alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT emilyengelhart alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT mirandaclahman alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT jeffreyadamo alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT xinyu alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT oliverserang alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT kylejminch alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT kimberlyywellman alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT davidayounger alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT randolphmlopez alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity AT ryanoemerson alphabindadomainspecificmodeltopredictandoptimizeantibodyantigenbindingaffinity |