New multi-modal foundation model advances functional antibody design

Researchers have developed the Antigen-specific Antibody Multi-modal Foundation Model (AAMFM), a novel model designed for functional antibody design. AAMFM learns unified representations of antibody sequences and structures by conditioning on antigen context, incorporating antigen information like geometric interfaces and epitope annotations. The model is further refined using Calibrated Direct Preference Optimization (Cal-DPO) to align learning with binding-specific objectives, demonstrating state-of-the-art performance in antigen-specific antibody engineering. AI

IMPACT This model advances the field of antibody engineering by improving the design of antigen-specific antibodies.

RANK_REASON The cluster contains a research paper detailing a new model and methodology. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• AAMFM
• alphaXiv
• Antigen-specific Antibody Multi-modal Foundation Model
• arXiv
• Cal-DPO
• Calibrated Direct Preference Optimization
• CatalyzeX
• DagsHub
• Gotit.pub
• Hugging Face
• ScienceCast

• paper
• model release

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