Neoantigens are special markers unique to cancer cells. Turning on the B-cell immune response allows cancer vaccines to go beyond short-term protection and create long-term immune memory that recognizes the tumor and reduces the risk of its return. Scientists from the Korean Institute of Advanced Technology have created a technology for personalized design of such vaccines using artificial intelligence, which allows them to select the most effective treatment for each patient.
A team of scientists from the Department of Bioengineering and Neuroengineering together with Neogen Logic Co., Ltd. developed a new AI model that predicts neoantigens, a key component of tailored cancer vaccines, and showed a significant role for B cells in cancer immunotherapy.
The researchers were able to overcome the limitations of existing methods that focused primarily on T-cell responses and created a neoantigen prediction system that takes into account both T- and B-cell responses. The technology was confirmed through large-scale analysis of tumor genomes, animal experiments and clinical trial data. It was recognized as the first AI model capable of quantifying B cell reactivity to neoantigens.
Neoantigens are formed from protein fragments that are altered due to mutations in cancer cells. Due to their high specificity, they are considered a promising target for new generation vaccines. mRNA platforms like those of Moderna and BioNTech have grown out of research in this area, and today the companies are actively testing personalized cancer vaccines with big pharma.
However, most current approaches still emphasize T-cell immunity and pay little attention to the role of B cells. A new AI model solves this problem: it analyzes how mutated proteins interact with B cell receptors (BCRs) and predicts their reactivity based on this. When analyzing clinical data from vaccines, researchers showed that the inclusion of a B-cell response significantly enhances antitumor protection in real practice.