KAIST Personalized Cancer Vaccine AI: World's First Model to Predict B-cell Immunity

Can AI finally make cancer recurrence a thing of the past? A joint research team from KAIST and biotechnology firm Neogenlogic has developed a new AI model that targets the very 'fingerprints' of tumors to create personalized vaccines.

KAIST Personalized Cancer Vaccine AI: Beyond T-cells

While current cancer vaccines focus almost exclusively on activating cytotoxic T-cells for immediate attacks, this new AI platform shifts the focus toward B-cell-mediated immune memory. It's widely believed that this memory is the secret to durable, long-term antitumor responses and preventing the illness from returning.

The model works by learning structural interaction patterns between mutant peptides and B-cell receptors (BCRs). According to the team led by Professor Choi Jung-kyoon, this is the world's first AI framework capable of predicting B-cell immunogenicity alongside T-cell responses for vaccine design. Their findings were published in the Dec. 3 edition of Science Advances.

Roadmap to FDA Approval and 2027 Trials

Neogenlogic has already integrated this AI into its proprietary discovery engine, DeepNeo. The technology has been validated against large-scale genomic datasets and clinical trial data from global vaccine leaders, ensuring its scientific rigor.

The team is now preparing an investigational new drug (IND) submission with the FDA. They plan to enter clinical trials in 2027. Professor Choi noted that the study provides the empirical evidence needed to move from theoretical prediction to systematic clinical application.