On fine-tuning Boltz-2 for protein-protein affinity prediction

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: King, James, Cornwall, Lewis, Nica, Andrei Cristian, Day, James, Sim, Aaron, Dalchau, Neil, Wollman, Lilly, Meyers, Joshua
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912753217699840
author King, James
Cornwall, Lewis
Nica, Andrei Cristian
Day, James
Sim, Aaron
Dalchau, Neil
Wollman, Lilly
Meyers, Joshua
author_facet King, James
Cornwall, Lewis
Nica, Andrei Cristian
Day, James
Sim, Aaron
Dalchau, Neil
Wollman, Lilly
Meyers, Joshua
contents Accurate prediction of protein-protein binding affinity is vital for understanding molecular interactions and designing therapeutics. We adapt Boltz-2, a state-of-the-art structure-based protein-ligand affinity predictor, for protein-protein affinity regression and evaluate it on two datasets, TCR3d and PPB-affinity. Despite high structural accuracy, Boltz-2-PPI underperforms relative to sequence-based alternatives in both small- and larger-scale data regimes. Combining embeddings from Boltz-2-PPI with sequence-based embeddings yields complementary improvements, particularly for weaker sequence models, suggesting different signals are learned by sequence- and structure-based models. Our results echo known biases associated with training with structural data and suggest that current structure-based representations are not primed for performant affinity prediction.
format Preprint
id arxiv_https___arxiv_org_abs_2512_06592
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle On fine-tuning Boltz-2 for protein-protein affinity prediction
King, James
Cornwall, Lewis
Nica, Andrei Cristian
Day, James
Sim, Aaron
Dalchau, Neil
Wollman, Lilly
Meyers, Joshua
Machine Learning
Biomolecules
Accurate prediction of protein-protein binding affinity is vital for understanding molecular interactions and designing therapeutics. We adapt Boltz-2, a state-of-the-art structure-based protein-ligand affinity predictor, for protein-protein affinity regression and evaluate it on two datasets, TCR3d and PPB-affinity. Despite high structural accuracy, Boltz-2-PPI underperforms relative to sequence-based alternatives in both small- and larger-scale data regimes. Combining embeddings from Boltz-2-PPI with sequence-based embeddings yields complementary improvements, particularly for weaker sequence models, suggesting different signals are learned by sequence- and structure-based models. Our results echo known biases associated with training with structural data and suggest that current structure-based representations are not primed for performant affinity prediction.
title On fine-tuning Boltz-2 for protein-protein affinity prediction
topic Machine Learning
Biomolecules
url https://arxiv.org/abs/2512.06592