On Jul. 24, 2025, University of Alberta (UA) researchers announced they have harnessed the power of artificial intelligence to design a blueprint for building a vaccine that aims to teach the body’s immune system to fight cancer.

“Our current focus is on addressing the growing challenges of melanoma, one of the most aggressive forms of skin cancer,” says Saba Ismail, a PhD student in co-author and pharmacy professor Khaled Barakat‘s lab in the Faculty of Pharmacy and Pharmaceutical Sciences and lead author of a paper introducing the vaccine, recently published in Computers in Biology and Medicine. “However, once we finalize this model, we will use it for other cancers as well, not just melanoma.” Although the vaccine is still in the theoretical phase (not yet tested on actual cells or in humans), it marks “an exciting step in a very long journey to come,” adds Barakat.

The computer model of the vaccine contains multiple neoantigens — markers found on cancer cells that identify them to the immune system as something foreign. The body’s T cells recognize these neoantigens and are triggered to eliminate them. Neoantigen-based vaccines are a promising approach for cancer immunotherapy, the authors note in the paper, but identifying which neoantigens should be targeted for different types of cancer, and in different patients, can be a challenge.

Computer modelling allows for a much faster and more efficient way to narrow down the potential neoantigens to a handful of promising targets. Ismail and Barakat began with 750 neoantigens and, with their model, found eight strong candidates they then combined into the vaccine “construct.”  The vaccine has shown promising results in computational testing. Ismail also notes it “has high binding affinity towards immune receptors,” another essential step for triggering the immune system.

The researchers stress that these findings are just the first step, highlighting the need for extensive testing in the lab followed by an eventual clinical trial.