A joint team from KAIST and Neogenlogic has developed the world’s first AI framework that predicts both T-cell and B-cell immunogenicity for neoantigens, enabling more effective personalized cancer vaccines focused on long-term immune memory and recurrence prevention.
Glimpse:
Published in Science Advances (December 2025), the AI model identifies patient-specific neoantigens that trigger robust B-cell responses critical for durable anti-tumor immunity beyond traditional T-cell-focused vaccines. Validated on large genomic datasets and clinical trials, it integrates into Neogenlogic’s DeepNeo platform, with preclinical work targeting FDA IND submission and trials by 2027.
South Korean researchers have unveiled a pioneering AI model that could redefine personalized cancer vaccines by incorporating B-cell immune responses, offering stronger, longer-lasting protection against tumor recurrence. The breakthrough comes from a collaboration between Professor Choi Jung-kyoon’s team at the Korea Advanced Institute of Science and Technology (KAIST) and biotechnology firm Neogenlogic.
Current cancer vaccines primarily activate cytotoxic T-cells for immediate tumor attacks, but emerging evidence shows B-cell-mediated immune memory is essential for sustained defense. The new AI framework addresses this gap: it learns structural interactions between mutant peptides (neoantigens) and B-cell receptors to predict which neoantigens will elicit powerful B-cell activation creating “immune memory” that prevents cancer return.
Key innovations
Dual prediction: World’s first to forecast both T-cell and B-cell immunogenicity.
High accuracy: Validated against massive genomic datasets and real-world clinical trial data from leading vaccine developers.
Integration: Fully embedded in Neogenlogic’s DeepNeo discovery engine for rapid neoantigen selection.
Neoantigens unique mutation-derived proteins on tumor cells serve as “fingerprints” for training the immune system. By prioritizing those sparking B-cell memory, the model promises vaccines with superior long-term efficacy.
The team is advancing preclinical development of a personalized vaccine platform, aiming for U.S. FDA Investigational New Drug (IND) submission and clinical trials in 2027.
This advancement positions South Korea at the forefront of AI-driven immuno-oncology, potentially transforming outcomes for patients facing recurrence risks.
“The study introduces the world's first AI framework capable of predicting B cell immunogenicity alongside T cell responses for the design of personalized cancer vaccines.”
By
HB Team
