ChatGPT, AlphaFold shrink dog tumor 75%

A self-taught ML engineer just ran a precision oncology pipeline that pharma companies spend billions on, for $3k and two months — and got results.

• –AlphaFold was the critical technical lever: it modeled the 3D structure of Rosie's tumor mutations to identify neoantigens, the specific protein targets most likely to trigger an immune response
• –ChatGPT served as an interdisciplinary research copilot — helping Conyngham navigate genomics and immunology with no formal biology training
• –UNSW's RNA Institute synthesized the actual mRNA vaccine from the AI-identified targets, underscoring that this workflow still requires academic wet-lab partners
• –The $3,000 total cost (DNA sequencing + vaccine synthesis + veterinary administration) is a landmark data point for the democratization argument around precision medicine
• –Moderna is already in human trials for personalized mRNA cancer vaccines using the same neoantigen logic — this story validates the approach and shows AI is compressing the design timeline dramatically