aiqbench analysis
This research applies Direct Preference Optimization (DPO) to protein language models for antibody design, effectively leveraging large-scale weak supervision from immunization data where labeled experimental results are scarce. By adapting DPO with union-masked log-likelihood approximations and IMGT-based alignment, the study demonstrates a scalable method that significantly outperforms baselines in optimizing expressibility across diverse camelid-derived sequences. The work highlights how advanced machine learning techniques can overcome critical data bottlenecks in computational biology to accelerate therapeutic protein development.
Preference-based Antibody Expression Ranking: Scaling with Large-scale Weak Supervision
arXiv:2607.16263v1 Announce Type: new Abstract: Antibody expression ranking is a critical task in antibody design, yet its modelling is severely hindered by the scarcity of labeled experimental data. To address this, we propose a unified preference-based learning framework that integrates scarce quantitative expression data with large-scale weak positive supervision from immunization data. We adapt Direct Preference Optimization (DPO) to protein language models by introducing a union-masked log-likelihood approximation and IMGT-based alignment, enabling efficient training on variable-length sequences. Evaluating on a diverse internal dataset of 1254 labeled sequences and 4 million unlabeled camelid-derived antibodies, we show that our method consistently outperforms baselines on most metrics. Our results demonstrate that preference learning can effectively learn from weak supervision, providing a scalable solution for antibody expressibility optimization in data-constrained settings. Project page: https://kisoji-biotechnology-inc.github.io/Preference-Expression-Ranking/.
Abstract:Antibody expression ranking is a critical task in antibody design, yet its modelling is severely hindered by the scarcity of labeled experimental data. To address this, we propose a unified preference-based learning framework that integrates scarce quantitative expression data with large-scale weak positive supervision from immunization data. We adapt Direct Preference Optimization (DPO) to protein language models by introducing a union-masked log-likelihood approximation and IMGT-based alignment, enabling efficient training on variable-length sequences. Evaluating on a diverse internal dataset of 1254 labeled sequences and 4 million unlabeled camelid-derived antibodies, we show that our method consistently outperforms baselines on most metrics. Our results demonstrate that preference learning can effectively learn from weak supervision, providing a scalable solution for antibody expressibility optimization in data-constrained settings. Project page: this https URL.
Machine Learning (cs.LG); Computational Engineering, Finance, and Science (cs.CE); Quantitative Methods (q-bio.QM)
arXiv:2607.16263 [cs.LG]
(or arXiv:2607.16263v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2607.16263 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 230 words
