Authors: Kate White/Lexi Luo

Paul Conyngham had no background in biology. He had a dying rescue dog, a data engineer’s instincts, and a year he refused to waste. 

Rosie is an eight-year-old Staffy cross. Paul rescued her in 2019. By May 2024, she had aggressive mast cell cancer spreading up and down both hind legs. In places, tumors had burst through the skin entirely. Three vets over 11 months had told him not to worry. When a biopsy was finally taken, the prognosis was months. 

What followed was one of the most remarkable acts of individual scientific determination in recent memory. While running his AI consulting business full time, Paul taught himself computational genomics from scratch. He had Rosie’s healthy and tumor DNA sequenced at UNSW. He used AlphaFold to model her mutated proteins, identified the c-KIT mutation driving the cancer, and narrowed the field to seven neoantigen targets her immune system could be trained to attack. He navigated ethics approvals across 2 universities, spent 120 hours on paperwork, and drove 14 hours to Queensland on Christmas Day 2025 so Rosie could receive her first injection. 

Six weeks later the tumors were visibly shrinking. By January 2026 she was jumping fences at the dog park chasing rabbits. 

It was the first personalized mRNA cancer vaccine ever designed for a dog, built by someone who had never studied biology in his life. The scientists who worked alongside him were stunned. When the story broke globally in March 2026, it accumulated millions of views within days. 

A Year That Should Have Been Months 

Read Paul’s account carefully, and two things stand out. The first is the brilliance of what he did. The second is how much of that year was not spent doing science at all. 

It was spent finding things out. 

When the cancer was diagnosed, Paul had no map. He started with ChatGPT, trying to understand what mast cell cancer and molecular biology were. Once he identified the c-KIT mutation, he had to figure out what that meant for treatment options, querying AI tools, cross-referencing papers, and slowly building a picture of a research landscape that already existed in full, somewhere, but was not immediately available to him. He ran simulations across a million compounds looking for a ligand that could bind to the mutated protein, and found one after two weeks. Then discovered it was patented. He tracked down the holder, asked for a compassionate use exemption, and was turned down. Weeks of computation, and a door he had no way of knowing was closed before he started. 

It felt, he said, like they had exhausted the space. He stepped back. He spent two weeks just being with Rosie. Then the vaccine idea came to him. He spent a night with ChatGPT debating possibilities, and eventually arrived at a direction nobody had pointed him toward. He had never heard of mRNA neoantigen vaccines before 2024. He spent months getting up to speed, through cold emails to researchers, more AI queries, more papers, more dead ends. It was through a chance connection with Dr. Mari Maeda of the Canine Cancer Alliance, who had been watching his work, that he finally found the right collaborators and institutions to make it possible. 

“I do not think everyone should have to know what I know to be able to use these tools. I think this was far more complex than it needed to be. The science exists. The AI exists. The gap is in making the right information reachable.” 
– Paul Conyngham 

What Paul Would Have Found 

Patsnap is a research intelligence platform used by innovation teams at the world’s leading pharmaceutical and biotech companies. It brings together patent landscapes, compound development data, clinical progress, researcher networks and more into one place. It is exactly the tool that could have changed Paul’s year.  

We ran these searches ourselves. Here is what Paul would have found.   

Paul’s first real breakthrough was identifying the c-KIT mutation. His first real setback was not knowing what to do with it. He spent weeks querying AI tools, reading papers, and slowly triangulating a treatment landscape that already existed in full, somewhere, just not in any form he could reach.  

A single question to Patsnap’s Hiro LS Researcher agent would have changed that.  

Within minutes it would have mapped the clinical significance of the mutation, connected it to known treatment strategies, and surfaced neoantigen vaccines as a viable pathway. Paul spent weeks getting there.  

This would have taken minutes. 

Exhibit 1: Summary of neoantigens & c-KIT mutation treatment strategies. Source: PatSnap Synapse.

From there, Patsnap Synapse would have shown him the full compound landscape for c-KIT. 157 candidates in clinical development. 17 already FDA-approved. Toceranib and masitinib right there in the results, both specifically developed for canine mast cell tumors with c-KIT mutations, IP status and clinical data clearly documented. Paul was working blind. He spent two weeks simulating a million compounds, found a candidate, and hit a patent wall he had no way of seeing coming. That door was always closed. Synapse would have told him before he walked toward it. 

Exhibit 2: Top 10 indications sorted by R&D status. Source: PatSnap Synapse.

After the patent dead end, Paul stepped back and spent two weeks with Rosie. The vaccine idea came to him in that space. He had never heard of mRNA neoantigen technology before 2024, and now he had to learn an entire field from scratch, alone, in the middle of everything else. A patent landscape search would have given him the map he was missing. Over 550 patent families filed in this space in the past five years alone. BioNTech, Moderna, Genentech at the frontier. Delivery mechanisms, combination approaches, the full architecture of how neoantigens get encoded into mRNA, all of it structured and searchable. 

Once he had this direction, Quick Research would have given him a read on the field itself. Sequence design, codon optimization, delivery models. Not scattered across papers he had to find and validate one by one, but consolidated, legible, ready to act on. Document Analyzer would have handled the validation work he found most grinding, extracting the most relevant findings across dense technical sources in a single session rather than days of manual cross-referencing. 

Exhibit 3: Research report on mRNA vaccine strategy. Source: PatSnap Eureka.

The last thing Paul needed, and the last thing he found, was the right people. It took months of cold emails before a chance connection with Dr. Mari Maeda of the Canine Cancer Alliance, who had been quietly watching his work, led him to Professor Rachel Allavena at the University of Queensland. The right collaborators were there all along. Inventor and assignee data in Patsnap returns a ranked map of the most active researchers in any field, by institution, by country, by recency. That connection would not have required months and luck. It would have required a search. 

Exhibit 4: Assignee analysis and relationships. Source: PatSnap Synapse.

None of this replaces what Paul did. The scientific instinct, the relentless execution, the researchers he brought together. Those were irreplaceable. But the information layer is exactly what Patsnap is built for. In a race against a terminal diagnosis, those months are not abstract. They are time with your dog. 

The information layer that was always missing 

For decades, the kind of research intelligence Paul needed has existed, but inside institutions. Pharma companies, biotech firms, and university research groups have always had access to structured patent landscapes, chemical databases, and expert networks. The difference is they had teams to navigate them. Paul had determination and an internet connection. 

That asymmetry is what Patsnap was built to close. Not by replacing the scientific instinct or the relentless execution, those were Paul’s. However, the information layer must be accessible to anyone who needs it. Knowing what already exists before you spend two weeks running simulations. Knowing which door is closed before you walk toward it. Knowing who to call. 

Paul’s story is extraordinary. But the situation he found himself in, a person with a real problem, real capability, and no clear map, is not. There will be more people like him. The question is how much of their time gets spent finding things out. 

Rosie is still running 

Paul is already working on a second vaccine for the tumor that did not respond to the first. The data came back in March. Analysis has begun. The story that started three years ago with a Staffy cross and some worrying lumps is still being written. 

What strikes you, reading Paul’s account from the beginning, is not just what he achieved. It is what it cost him to get there. The months spent learning fields from scratch. The cold emails. The dead ends. The two weeks sitting with Rosie after a failure, waiting for an idea that might not come. He did all of it out of love for one dog, and he did it the hardest possible way. 

The science did not have to be that hard. The information was always out there. It just was not reachable. That is the thing worth fixing, and it is exactly what Patsnap is for.