The Final Frontier for Ending Cancer

What does FinalDose do?

FinalDose is building the final frontier for ending cancer. Industry targets proteins, but most cancer-driving mutations produce proteins indistinguishable from healthy ones. FinalDose is the only startup using nucleic acid, DNA and RNA, to search and destroy cancer cells: infinite differentiability and programmability, the last modality to end all cancers.

The current landscape

For 60 years, drug discovery has chased proteins. Pick a protein, screen millions of molecules against it, hope something sticks. Ten years and $2.6B per drug. But ~85% of disease drivers don't have a druggable protein, making them unreachable for modern medicine. The deeper issue: most disease doesn't start at the protein, it starts in the DNA that codes for it. The protein is the symptom, not the cause.

Every modality in oncology today is a different way of reading a protein, and each one inherits the same ceiling.

Small molecules need a pocket to bind. Most cancer-driving mutations don't make one. MYC, TP53 and APC have been known as drivers for forty years and are still called undruggable. Where small molecules do work, they modulate rather than eliminate, so the pattern is response, resistance, progression.

Antibody-drug conjugates need a surface protein that healthy cells don't express. There are only a few hundred candidates, the good ones are taken, and the payloads are chemotherapy molecules first described between 1972 and 1987. The targeting got smarter. The killing never did.

CAR-T needs the same clean surface marker, and then borrows the immune system's own kill chemistry, granzyme and perforin, which is 500 million years old. It has been transformative in blood cancers and has largely failed to reach solid tumors.

Checkpoint inhibitors don't target the cancer at all. They release a brake and hope the immune system finds it.

All four share one dependency: the difference between a cancer cell and a healthy cell has to show up at the protein surface. Most of the time it doesn't. The mutation is in the DNA, and the protein it produces looks the same as everyone else's.

So FinalDose went straight for the DNA. FinalDose built a programmable drug that reads nucleic acid inside the cell, decides if it's diseased, and destroys the cell if it is. One platform, reprogrammed to any disease distinguishable on a DNA level. Same molecule, new guide, new drug.

A handful of companies are now asking a version of the same question, and they're the real

comparison set: Trogenix, which gates a kill payload on a cancer-state transcriptional switch and dosed its first glioma patient in May 2026 off a £70M Series A; Earli and Strand Therapeutics, both gating on cell state rather than genotype.

None of them read the genome itself and the company believes it is the final frontier. That's the piece FinalDose owns.

How FinalDose started

FinalDose started with three Oxford PhDs nerding out.

We were going through the programmable systems biology has evolved, we found something fascinating with one of the systems. When it finds its target, it undergoes a dramatic structural change. The whole protein physically reconfigures.

That motion is normally treated as a step on the way to editing a gene. We looked at it and had an idea: why can't we wire that reconfiguration into a kill switch and destroy the cell?

Nobody had tried it. That's what got us excited because of what it implies. It's a completely new level of resolution in biology for removing unwanted cells. Not by what they look like on the outside, but by what they are written as on the inside. This literally confers near boundary-limiting resolution!

And once you can do that, cancer is the first application, not the whole idea. Think about what the body is actually doing over time. It is continuously breaking down and producing unwanted cells, and their slow accumulation is what degrades the entire organism: cancer, chronic disease, aging.

So if we crack this, we don't just get a cancer drug. We get to go after the slow decay of biology itself.

We filed the foundational patent at the end of April 2026 and proved the switch worked in the lab six weeks later.

The FinalDose solution

Our drug is one engineered protein, reprogrammed for any cancer.

It does three things. It carries a guide and searches for one sequence. When it finds it, the

protein physically changes shape, and that change flips a switch we built into it. The switch

releases a payload that kills the cell. In a healthy cell, if no match, there is no release, and the cells are untouched. We are not editing the genome. We read it and decide whether that cell should exist.

The switch is the invention, and it doesn't care what sequence you point it at. Change the guide, you have a new drug. Same molecule, same manufacturing, new disease. That's what makes this a platform and not a drug.

It also reaches things protein-based drugs can't see at all: mutations in DNA that never becomes a protein, dormant virus sitting in the genome, fusion points that exist in zero healthy cells. And we don't have to discover any of it. Twenty years of cancer genome sequencing already tells us which sequences cause which disease.

Cancer is where we start, not where we stop. Every disease we go after is the same

molecule with a different guide. And it doesn't have to be cancer. Anything that comes down to unwanted cells with a DNA/RNA address.

Delivery is the hard part, and it's something that needs to be solved, whether its by the community or by us. Once it is, we have a general way to remove any cell you can name.

Cancer is the first application, not the whole idea. Our body is constantly producing cells we don't want, and their slow accumulation is what breaks the whole organism down. If we get this right, we're not building a cancer drug. We're going after the slow decay of biology itself.

>

— Jeff Liu, The Dude at FinalDose

The team culture at FinalDose

Built on the three principles from your operating system, with one real anecdote.

FinalDose is trying to end cancer. Nobody on this team pretends that is a normal job, so the team does not run the company like one.

Three things the company runs by:

Maximum ownership. One name on every single experiment and action. Never a team and this is rare in biology.

Biology is the only allowed friction: Cells take three days to grow. That is

physics, and the company accepts it. Waiting on a decision, a reagent, a booking, an invoice or a meeting is not biology and there should be close to no friction on those actions.

Attached to the mission, not the asset: In biology, it is very easy to fall in love with the company's asset (or drug you're working on). At some point the company will kill something someone on the team worked on for months. This is not a failure! This is FinalDose figuring out the shortest critical path has shifted.