Want another example of a critical way station on the path to genomic cures with AI? One inspiring story comes from Australia, where a tech entrepreneur named Paul Conyngham, after seeing his dog, Rosie, develop leg tumors, decided to take matters into his own hands using GPT o1, Gemini 2, and Grok 3.
First, some context: with models like Claude Operon popping onto the radar, and biosciences companies leveraging the human knowledge of the genome, we're a good way beyond Alphafold, and venturing into territory where CRISPR-esque tools will allow humanity to cure disease and optimize health in profound, fundamental ways.
So it may be no surprise to those who understand what AI can do on this front that with the right plan and the right tools, one can take available data, sequence DNA, and build interventions based directly on protein and receptor binding. This is similar to what new drugs often do, targeting specific pathways for very specific tweaks to human (or animal) anatomical processes.
The Timeline
For the interested, we're lucky to have a detailed narrative straight from Conyngham himself, in the form of a multi-page X post where the innovator describes his journey toward curing his dog's cancer. The symptoms first appeared in June of 2023, but it took until May of 2024, after various vets didn't do much, that Conyngham decided to embark on a new treatment plan.
By spring of 2025, the involved team had isolated the problem and begun to model something called a "mutated c-KIT protein," and were thinking about using a ligand to help. Conyngham provides this definition of ligand, helpfully, in the margin:
Any molecule that binds specifically to a larger "target" molecule (usually a protein) to serve a purpose
"Pulling together more scripts and simulation engines, we developed a candidate that worked in simulation," Conyngham explains. "It could have taken years to go from petri dishes, to mouse models and finally to dogs. After further consideration, the risks and approvals process for a novel ligand made this approach unworkable within Rosie's timeframe. But the ligand candidate still exists - and if the approvals landscape ever catches up, it may yet have its day."
By that fall, Conyngham was navigating ethics approval. By around Christmas, he was ready to try an mRNA vaccine on Rosie.
By February, tumor growth on the legs had halted. Another tumor was removed and sent for analysis.
So this story is still ongoing, but it exemplifies the kinds of things we are likely to see in modern medicine going forward, where AI really makes the impossible possible. It sounds trite, but I think that's really what's happening.
The Stakeholders
Reading through Conyngham's testimony, you can see that various health offices were involved in the project. Toward the end, Conyngham makes an effort to explain that the chatbots only did certain things related to the research, and the humans did the rest: specifically, he names Professor Pall Thordarson at the UNSW mRNA Institute who manufactured the vaccine; Professor Rachel Allavena & Dr. José Granados at the University of Queensland who administered it; and Professor Martin Smith at UNSW Sydney.
"Here's what the chat bots did NOT do," Conyngham writes. "They did not collect samples. They did not isolate or sequence the DNA. They did not physically manufacture the vaccine. They did not administer it."
Instead, the models were involved in planning and pipeline design, troubleshooting, candidate filtering (w/ Grok) and treatment protocol design. In some ways, that's actually a lot, but it's not "automation," per se. It's collaborative.
After using a "Star Trek" metaphor to try to explain tyrosine kinase and PD-1 inhibitors, Conyngham summarizes his idea of how AI works in this process:
"The chat bots empowered me as an individual to act with the power of a research institute - planning, education, troubleshooting, compliance, and yes, real scientific design work in converting genomic data to a vaccine prescription and designing the treatment protocol around it. But they worked alongside humans at every step. The combination is what made it possible."
He also contrasts the promise and possibility of AI-driven cancer research with all of the obstacles and gatekeeping that hold progress back.
"I do not know why I had to be this obsessed to get here," Conyngham writes. "There are so many unnecessary barriers. So many techniques and tools that could be made far easier to access. The science exists. The AI exists. The gap is in making it reachable."
To be clear, though Rosie is showing signs of improvement, this isn't the end of the tale. But again, it's inspiring, especially in context. If one person can do this for one dog, can't we make quick strides in wiping out cancer or at least proliferating treatments? And won't health look pretty different pretty soon?