Saved in:
Bibliographic Details
Main Authors: Chen, Can, Herpoldt, Karla-Luise, Zhao, Chenchao, Wang, Zichen, Collins, Marcus, Shang, Shang, Benson, Ron
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.10365
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915153812914176
author Chen, Can
Herpoldt, Karla-Luise
Zhao, Chenchao
Wang, Zichen
Collins, Marcus
Shang, Shang
Benson, Ron
author_facet Chen, Can
Herpoldt, Karla-Luise
Zhao, Chenchao
Wang, Zichen
Collins, Marcus
Shang, Shang
Benson, Ron
contents Antibodies are widely used as therapeutics, but their development requires costly affinity maturation, involving iterative mutations to enhance binding affinity.This paper explores a sequence-only scenario for affinity maturation, using solely antibody and antigen sequences. Recently AlphaFlow wraps AlphaFold within flow matching to generate diverse protein structures, enabling a sequence-conditioned generative model of structure. Building on this, we propose an alternating optimization framework that (1) fixes the sequence to guide structure generation toward high binding affinity using a structure-based affinity predictor, then (2) applies inverse folding to create sequence mutations, refined by a sequence-based affinity predictor for post selection. A key challenge is the lack of labeled data for training both predictors. To address this, we develop a co-teaching module that incorporates valuable information from noisy biophysical energies into predictor refinement. The sequence-based predictor selects consensus samples to teach the structure-based predictor, and vice versa. Our method, AffinityFlow, achieves state-of-the-art performance in affinity maturation experiments. We plan to open-source our code after acceptance.
format Preprint
id arxiv_https___arxiv_org_abs_2502_10365
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle AffinityFlow: Guided Flows for Antibody Affinity Maturation
Chen, Can
Herpoldt, Karla-Luise
Zhao, Chenchao
Wang, Zichen
Collins, Marcus
Shang, Shang
Benson, Ron
Machine Learning
Antibodies are widely used as therapeutics, but their development requires costly affinity maturation, involving iterative mutations to enhance binding affinity.This paper explores a sequence-only scenario for affinity maturation, using solely antibody and antigen sequences. Recently AlphaFlow wraps AlphaFold within flow matching to generate diverse protein structures, enabling a sequence-conditioned generative model of structure. Building on this, we propose an alternating optimization framework that (1) fixes the sequence to guide structure generation toward high binding affinity using a structure-based affinity predictor, then (2) applies inverse folding to create sequence mutations, refined by a sequence-based affinity predictor for post selection. A key challenge is the lack of labeled data for training both predictors. To address this, we develop a co-teaching module that incorporates valuable information from noisy biophysical energies into predictor refinement. The sequence-based predictor selects consensus samples to teach the structure-based predictor, and vice versa. Our method, AffinityFlow, achieves state-of-the-art performance in affinity maturation experiments. We plan to open-source our code after acceptance.
title AffinityFlow: Guided Flows for Antibody Affinity Maturation
topic Machine Learning
url https://arxiv.org/abs/2502.10365