Light Horse Therapeutics

A startup that walked out of stealth already holding a billion-dollar deal

On a January morning in 2025, a company most of the industry had never heard of made two announcements at once. The first: $62 million in Series A funding. The second, arguably louder: a research collaboration with Novartis worth up to $1 billion in milestones, signed before the public knew the company existed. Light Horse Therapeutics did not arrive quietly.

Today Light Horse is a roughly 38-person biotech operating out of San Diego's Nancy Ridge biotech corridor. It is discovery-stage, which in biotech means it has no product on a shelf and no patient on a clinical trial - yet. What it has instead is a platform and a thesis, and pharma's biggest names have already paid to bet on both.

The pitch is deceptively simple, which is the most dangerous kind of simple. Light Horse believes the standard way of discovering drugs has the steps in the wrong order. So it reversed them.

Drug hunters screen millions of molecules and hope. The protein never gets a say.

Conventional small-molecule discovery works roughly like this: pick a protein you think matters in a disease, find a pocket on its surface, then screen enormous chemical libraries hoping something sticks well enough to do something useful. The trouble is that a molecule can bind a protein beautifully and still change nothing. Binding is not the same as breaking the thing that drives the disease.

This is why so much of biology has earned the unflattering label "undruggable." Many of the proteins most central to cancer have smooth surfaces, no obvious pockets, or sites that, when hit, simply don't matter. Decades of effort have been spent making molecules that grip proteins in the wrong places.

The question Light Horse's founders kept circling was uncomfortable: what if the field was screening for chemistry before it actually knew which part of the protein was worth attacking? What if you could ask the protein first?

The wall every cancer chemist hits. Roughly the proteins most worth drugging are the ones least willing to be drugged. Light Horse decided that was a feature to exploit, not a wall to accept.

Three chemical biologists, one inverted idea

Light Horse was built inside Versant Ventures' Inception Discovery Engine and pulled together three scientific co-founders whose names carry weight in chemical biology: Ben Cravatt of Scripps Research, Nathanael Gray of Stanford, and Brian Liau of Harvard. Between them they cover chemoproteomics, medicinal chemistry, and CRISPR-based functional genomics - which, conveniently, are exactly the three ingredients the company's platform needs.

Their bet was to combine systematic CRISPR-based gene editing with chemistry to do something the field rarely does in order: first identify which exact regions of a protein are functionally critical for driving disease, and only then go hunting for small molecules that target those precise sites. Function first. Chemistry second.

A platform that interrogates proteins where they actually live

The core of Light Horse is a proprietary "function-first" platform. It uses systematic CRISPR-based gene editing to scan a disease-critical protein and reveal which of its regions are functionally essential - the parts you cannot disturb without breaking the protein's role in the disease. Crucially, it does this with the protein in its native cellular environment, not isolated in a test tube where its real behavior can vanish.

Once those functional sites are mapped, the company's chemists go looking for small molecules that bind those exact spots. The result, in theory, is a drug candidate aimed not at a convenient pocket but at the protein's actual weak point.

The first focus is oncology - specifically the high-value, historically stubborn cancer targets that have resisted conventional approaches. The company has signaled it could extend the same platform to other therapeutic areas later. For now, cancer is where the thesis gets tested hardest.

The map before the road. CRISPR draws the map of where a protein is vulnerable. Medicinal chemistry builds the road to get there. Doing it in that order is the whole company.

A short history of a young company

When skeptical money writes a check, pay attention

A discovery-stage biotech has no revenue and no approved drug to point to. So the proof, for now, is in who chose to fund it. The Series A was led by Versant Ventures, the firm that built the company, but the more telling signatures belong to the strategic investors: Bristol-Myers Squibb, AbbVie, and Taiho Ventures. These are pharmaceutical companies that evaluate drug-discovery platforms for a living, and they put money in.

Then there is Novartis, which agreed to a multi-target collaboration and a $25 million upfront payment - real cash, before any drug exists - against the possibility of up to $1 billion in milestones. That is not a courtesy. That is a large pharma deciding the platform is worth building programs on.

The backers are quotable, too. "We believe the talented team at Light Horse and its technology will enable a pipeline of first-in-class drug candidates," said Rami Hannoush, a Versant venture partner and board member - the kind of sentence investors say when they mean it and when they're hoping.

Two words on the wall: patients can't wait

Light Horse describes itself as dedicated to tackling the toughest challenges in medicine - developing transformative therapies for serious diseases through bold, cutting-edge science. Its stated values run through the usual biotech vocabulary: innovation, collaboration, scientific integrity, transparency, trust, data-driven decisions. The one that does more work than the others is "a sense of urgency," distilled into the company's guiding line: patients can't wait.

It is an easy phrase to print and a hard one to live by, because the function-first approach is, by design, slower up front. Mapping a protein's functional sites before screening chemistry is more deliberate than the spray-and-pray alternative. The bet is that doing the patient work early means fewer dead ends later - that the long way around is actually the short way to a drug that matters.

Back to that January morning

Return to the launch. Two announcements, one company nobody had heard of, and an industry that briefly looked up from its coffee. What made the debut interesting was not the dollar figures, impressive as they were. It was what the dollar figures implied: that serious, skeptical pharma money had looked at the idea of asking a protein where it breaks before drugging it, and decided the idea was worth real money.

The hard part is still ahead. Platforms are promises, and biology has humbled better-funded ones. Light Horse has no approved drug, no clinical data yet, and the same long odds that face every discovery-stage company trying to drug the targets everyone else abandoned. None of the money guarantees a molecule that helps a single patient.

But the morning it launched, Light Horse changed one small thing about how a stubborn problem gets approached. It said, out loud and with a billion-dollar partner standing next to it, that the order of the steps matters - that you should find the weak point first and build the molecule second. Whether that conviction becomes a cancer drug is the only question that ultimately counts. For a company whose motto is that patients can't wait, it is also the only deadline that does.

The company named after fast cavalry. Light horse units were prized for speed and reaching places heavier forces couldn't. The metaphor is a little on the nose. It also happens to fit.