A new AI-driven platform named MAGE (Monoclonal Antibody Generator) uses machine learning on antibody sequences to rapidly design entirely new antibodies against viral proteins, removing reliance on existing templates.

Developed at Vanderbilt University Medical Center, MAGE advances protein language modeling to enable quick design of antibodies in response to viral threats.

AI-driven tools like MAGE could transform immunology and therapeutic development by enabling real-time design and deployment of biologics for diseases lacking effective treatments, signaling a move toward rapid, precision medicine.

This work is a broad, multi-institution collaboration across the United States, Australia, and Sweden, emphasizing the fusion of computational methods with experimental biology.

Funding comes from ARPA-H and the NIH, underscoring sustained investment in AI-enabled biomedicine and its potential to improve public health responses.

Bypassing the need for prior samples or templates, MAGE could accelerate therapeutic development during outbreaks and broaden applications to cancer and autoimmune diseases.

As a proof of concept, MAGE generated antibodies that recognize unique H5N1 influenza antigen sequences, suggesting potential extrapolation to unknown strains and other viruses such as RSV and avian influenza.