New preference-based learning framework enhances antibody design

Researchers have developed a novel preference-based learning framework to improve antibody expression ranking, a crucial step in antibody design. This method leverages scarce quantitative expression data alongside a large dataset of weak positive supervision from immunization data. By adapting Direct Preference Optimization (DPO) for protein language models and incorporating IMGT-based alignment, the framework efficiently trains on variable-length sequences. Evaluations on a substantial internal dataset demonstrated that this approach consistently surpasses existing baselines, offering a scalable solution for optimizing antibody expressibility in data-limited scenarios. AI

IMPACT This research could accelerate the development of new therapeutics by improving the efficiency of antibody design.

RANK_REASON The cluster contains an academic paper detailing a new method for a specific scientific task. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• Antibody Expression Ranking
• Direct Preference Optimization
• Hugging Face
• IMGT/LIGM-DB
• protein language models

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →