A KAIST and Neogenlogic team has developed the first AI model to predict B and T cell responses for customised cancer vaccine design.
SEOUL: In a groundbreaking development, a joint team from the Korea Advanced Institute of Science and Technology (KAIST) and Neogenlogic, a South Korean biotech firm, has unveiled an artificial intelligence (AI) model capable of designing personalised cancer vaccines. This innovation is poised to transform how oncology treatments are tailored for individual patients.
At the core of the new system is the identification of neoantigens — unique protein fragments derived from mutations found in a patient’s tumour. These “fingerprints” serve as targets for training the immune system through vaccination, helping to prevent cancer recurrence.
What sets this model apart is its dual focus. While existing cancer vaccines primarily target cytotoxic T cells, which deliver immediate responses, the new AI framework also incorporates B cell-mediated immune memory, considered crucial for long-term tumour resistance.
“Emerging clinical data suggest that B cell responses offer durable protection,” the research team noted. Their model predicts which neoantigens are most likely to stimulate a robust B cell response, using learned interaction patterns between mutant peptides and B cell receptors (BCRs).
This is the first AI system in the world capable of evaluating both B cell immunogenicity and T cell reactivity, marking a significant milestone in immunotherapy and cancer vaccine design.
The new framework could soon guide pharmaceutical companies in creating more effective, long-lasting treatments tailored to the genetic profile of individual cancer patients, a leap forward in personalised medicine.
