Machine learning optimization of candidate antibody yields highly diverse sub-nanomolar affinity antibody libraries
Abstract Therapeutic antibodies are an important and rapidly growing drug modality. However, the design and discovery of early-stage antibody therapeutics remain a time and cost-intensive endeavor. Here we present an end-to-end Bayesian, language model-based method for designing large and diverse li...
Main Authors: | , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Nature Portfolio
2023-06-01
|
Series: | Nature Communications |
Online Access: | https://doi.org/10.1038/s41467-023-39022-2 |
_version_ | 1797801299170820096 |
---|---|
author | Lin Li Esther Gupta John Spaeth Leslie Shing Rafael Jaimes Emily Engelhart Randolph Lopez Rajmonda S. Caceres Tristan Bepler Matthew E. Walsh |
author_facet | Lin Li Esther Gupta John Spaeth Leslie Shing Rafael Jaimes Emily Engelhart Randolph Lopez Rajmonda S. Caceres Tristan Bepler Matthew E. Walsh |
author_sort | Lin Li |
collection | DOAJ |
description | Abstract Therapeutic antibodies are an important and rapidly growing drug modality. However, the design and discovery of early-stage antibody therapeutics remain a time and cost-intensive endeavor. Here we present an end-to-end Bayesian, language model-based method for designing large and diverse libraries of high-affinity single-chain variable fragments (scFvs) that are then empirically measured. In a head-to-head comparison with a directed evolution approach, we show that the best scFv generated from our method represents a 28.7-fold improvement in binding over the best scFv from the directed evolution. Additionally, 99% of designed scFvs in our most successful library are improvements over the initial candidate scFv. By comparing a library’s predicted success to actual measurements, we demonstrate our method’s ability to explore tradeoffs between library success and diversity. Results of our work highlight the significant impact machine learning models can have on scFv development. We expect our method to be broadly applicable and provide value to other protein engineering tasks. |
first_indexed | 2024-03-13T04:48:20Z |
format | Article |
id | doaj.art-4b310d77320b4b48bba9cd5498b95743 |
institution | Directory Open Access Journal |
issn | 2041-1723 |
language | English |
last_indexed | 2024-03-13T04:48:20Z |
publishDate | 2023-06-01 |
publisher | Nature Portfolio |
record_format | Article |
series | Nature Communications |
spelling | doaj.art-4b310d77320b4b48bba9cd5498b957432023-06-18T11:19:38ZengNature PortfolioNature Communications2041-17232023-06-0114111210.1038/s41467-023-39022-2Machine learning optimization of candidate antibody yields highly diverse sub-nanomolar affinity antibody librariesLin Li0Esther Gupta1John Spaeth2Leslie Shing3Rafael Jaimes4Emily Engelhart5Randolph Lopez6Rajmonda S. Caceres7Tristan Bepler8Matthew E. Walsh9Massachusetts Institute of Technology Lincoln LaboratoryMassachusetts Institute of Technology Lincoln LaboratoryMassachusetts Institute of Technology Lincoln LaboratoryMassachusetts Institute of Technology Lincoln LaboratoryMassachusetts Institute of Technology Lincoln LaboratoryA-Alpha Bio, Inc.A-Alpha Bio, Inc.Massachusetts Institute of Technology Lincoln LaboratoryResearch Laboratory of Electronics, Massachusetts Institute of TechnologyMassachusetts Institute of Technology Lincoln LaboratoryAbstract Therapeutic antibodies are an important and rapidly growing drug modality. However, the design and discovery of early-stage antibody therapeutics remain a time and cost-intensive endeavor. Here we present an end-to-end Bayesian, language model-based method for designing large and diverse libraries of high-affinity single-chain variable fragments (scFvs) that are then empirically measured. In a head-to-head comparison with a directed evolution approach, we show that the best scFv generated from our method represents a 28.7-fold improvement in binding over the best scFv from the directed evolution. Additionally, 99% of designed scFvs in our most successful library are improvements over the initial candidate scFv. By comparing a library’s predicted success to actual measurements, we demonstrate our method’s ability to explore tradeoffs between library success and diversity. Results of our work highlight the significant impact machine learning models can have on scFv development. We expect our method to be broadly applicable and provide value to other protein engineering tasks.https://doi.org/10.1038/s41467-023-39022-2 |
spellingShingle | Lin Li Esther Gupta John Spaeth Leslie Shing Rafael Jaimes Emily Engelhart Randolph Lopez Rajmonda S. Caceres Tristan Bepler Matthew E. Walsh Machine learning optimization of candidate antibody yields highly diverse sub-nanomolar affinity antibody libraries Nature Communications |
title | Machine learning optimization of candidate antibody yields highly diverse sub-nanomolar affinity antibody libraries |
title_full | Machine learning optimization of candidate antibody yields highly diverse sub-nanomolar affinity antibody libraries |
title_fullStr | Machine learning optimization of candidate antibody yields highly diverse sub-nanomolar affinity antibody libraries |
title_full_unstemmed | Machine learning optimization of candidate antibody yields highly diverse sub-nanomolar affinity antibody libraries |
title_short | Machine learning optimization of candidate antibody yields highly diverse sub-nanomolar affinity antibody libraries |
title_sort | machine learning optimization of candidate antibody yields highly diverse sub nanomolar affinity antibody libraries |
url | https://doi.org/10.1038/s41467-023-39022-2 |
work_keys_str_mv | AT linli machinelearningoptimizationofcandidateantibodyyieldshighlydiversesubnanomolaraffinityantibodylibraries AT esthergupta machinelearningoptimizationofcandidateantibodyyieldshighlydiversesubnanomolaraffinityantibodylibraries AT johnspaeth machinelearningoptimizationofcandidateantibodyyieldshighlydiversesubnanomolaraffinityantibodylibraries AT leslieshing machinelearningoptimizationofcandidateantibodyyieldshighlydiversesubnanomolaraffinityantibodylibraries AT rafaeljaimes machinelearningoptimizationofcandidateantibodyyieldshighlydiversesubnanomolaraffinityantibodylibraries AT emilyengelhart machinelearningoptimizationofcandidateantibodyyieldshighlydiversesubnanomolaraffinityantibodylibraries AT randolphlopez machinelearningoptimizationofcandidateantibodyyieldshighlydiversesubnanomolaraffinityantibodylibraries AT rajmondascaceres machinelearningoptimizationofcandidateantibodyyieldshighlydiversesubnanomolaraffinityantibodylibraries AT tristanbepler machinelearningoptimizationofcandidateantibodyyieldshighlydiversesubnanomolaraffinityantibodylibraries AT matthewewalsh machinelearningoptimizationofcandidateantibodyyieldshighlydiversesubnanomolaraffinityantibodylibraries |