Imperagen Says Its Enzymes Are 677x Better. The Question Is Whether the Quantum Label Holds Up.

A Manchester spin-out says its enzymes are 677 times better than what nature made. The question is why.

Imperagen, a University of Manchester spin-out founded in November 2021, closed a £5 million seed round on May 21, 2026, led by PXN Ventures with participation from IQ Capital and Northern Gritstone. The company says it engineered two enzymes 677 times and 572 times more productive than their natural counterparts in just five rounds of directed evolution — work done for an undisclosed Fortune 500 personal care company. Total funding to date sits at £8.5 million, including prior grant and angel backing.

That is a real number from a real client engagement. What is less clear is the mechanism. Imperagen describes its platform as a closed loop combining quantum physics simulations, problem-specific AI models, and automated robotics — a stack the company says can explore millions of mutations in silico before the lab even opens. Guy Levy-Yurista, who joined as CEO and has two exits across the US and Europe, frames it as a departure from the slow, manual enzyme engineering that pharmaceutical, personal care, and fine chemicals companies have lived with for decades.

But here is the part the press release does not linger on: the word "quantum."

The enzyme engineering world is crowded with companies claiming computational advantages. Cradle Bio, a direct competitor, describes its own platform as combining machine learning and automated lab execution — without the quantum label. Other players like Recursion Pharmaceuticals and Healx apply similar closed-loop AI-plus-robotics approaches to drug and rare disease discovery respectively. What Imperagen adds, according to its company website, is a layer of what it calls "QM Simulations" — quantum mechanics calculations that model how enzyme mutations will behave at the atomic level, before testing them in the lab.

The question is whether that layer is novel hardware or just a relabeling of methods that every computational biochemistry group has used for years. Density functional theory (DFT) and hybrid quantum mechanics/molecular mechanics (QM/MM) approaches are standard tools in academic enzyme design. They are not quantum computing in any hardware sense. They are physics-based simulations run on classical computers — useful, well-established, and not unique to Imperagen.

"We are not claiming quantum advantage in the hardware sense," a company spokesperson told TechCrunch, according to its coverage of the announcement. What the company means by "quantum physics" is the application of quantum mechanical equations to enzyme modeling. That is a defensible description of standard computational chemistry — and it is the same description any academic protein engineer would use for their DFT work.

This matters for investors and potential partners evaluating whether Imperagen has a genuine edge or is dressing a well-known method in fashionable language. The 677x improvement is the more concrete claim, and it comes from a client engagement rather than a published benchmark. One Fortune 500 client. One undisclosed application. No independent verification.

Dr. Andrew Currin, Dr. Tim Eyes, and Dr. Andy Almond — the three Manchester Institute of Biotechnology researchers who founded the company in November 2021 — have published in computational enzyme design and directed evolution. A search of their academic profiles shows no patents or publications directly describing novel quantum computing hardware or a proprietary quantum advantage claim. Their academic work appears to focus on the simulation methods that are now part of Imperagen's stack.

The distinction matters beyond the marketing. If the 677x result comes from the closed-loop automation — fast iteration, AI-guided selection of variants, robotic execution — then that is the real story, and it belongs to the broader category of AI-accelerated biology that Cradle Bio, Generate:Biomedicines, and a dozen other companies are also building. Quantum mechanics as a simulation layer may be genuinely useful for narrowing the search space before the lab. But that is a computational chemistry contribution, not a quantum computing one.

What Imperagen has that competitors often lack is a commercial validation. The client engagement produced a measurable result that the company can point to, which is more than many AI-biotech startups can claim at seed stage. The target sectors — pharmaceuticals, life sciences, personal care, sustainable fine chemicals, and industrial biotech — are real and diverse. Levy-Yurista's operational experience is real. The Manchester Institute pedigree is real.

The quantum label may turn out to be accurate in a narrow computational chemistry sense. Or it may be a way of signaling to VC audiences that Imperagen is operating at a more fundamental physical layer than its competitors. Those are not the same thing, and the press release does not resolve which one it intends. For now, the 677x number is the more honest anchor — a closed-loop, AI-guided, robotics-assisted engineering result that happened to use quantum mechanical simulations as one input among several. Whether that makes Imperagen a quantum biotech company or simply a well-instrumented one is a question the company has left for investors to answer themselves.