It should be noted right away that Coningham is not an amateur who used a “magic button” but a Data Science expert with 17 years of experience.
AI did not create a vaccine, it only helped to
In 2024, Rosie was diagnosed with mast cell cancer – a fairly common form of cancer in dogs and virtually untreatable. Surgery and chemotherapy halted the progression of the disease, but the dog’s condition continued to deteriorate – her coat became faded and her mobility was greatly reduced.
Paul Coningham decided to consult ChatGPT, directing the algorithm to immunotherapy and pointing to the Center for Genomics at the University of New South Wales (UNSW). The data analyst used AlphaFold, an AI program developed by Google DeepMind, to predict the 3D structures of mutated proteins and proprietary ML algorithms to select neoantigens.
He passed some of the data to Professor Pall Thorgdarson of the RNA Institute at UNSW, and the latter synthesized the nanoparticle. The vaccine was administered to the dog in December 2024 by Professor Rachel Allavena from the University of Queensland (Australia), who is the only veterinary researcher in the country with a valid ethical approval for such experiments. As early as January 2025, the tumor had halved, Rosie’s coat was shiny and her activity increased – she began jumping fences again.
ChatGPT was the research assistant, and critical decisions required the expertise of Prof. Thorgdarson’s team. AI accelerated the synthesis process that would have taken many months of manual literature review.
Human diseases are much more difficult to deal with
Experimental procedures in veterinary medicine operate under rules that are fundamentally different from human medicine. Testing is much faster, and so it can take less than two months from vaccine design to the first injection, as in the case of Coningham and his dog. This isn’t because AI has sped up the process – it takes many years to get approvals to make medicines for humans.
Despite this, history proves the fact that creating personalized mRNA vaccines against cancer is not the stuff of science fiction. According to Professor Thorgdarson, the use of AI democratizes the entire process. “What is amazing is that a Data-engineer with no biological background was able to generate a mRNA vaccine recipe,” he emphasized.
