A printer-sized machine helped. Not a supercomputer the size of a football stadium. A quantum box built by a British startup named ORCA. The Tech University of Denmark (DTU) plugged it into their AI models. It worked. Better, even, than classical methods for certain tasks.
They did it on weekends. They used leftover cash. Most science funding is rigid. Risky ideas die in grant applications. So, they paid out of pocket.
“Most innovative science is too scary for foundations.” — Timothy Patrick Jenkins, DTU professor and project lead.
The goal was simple: generate new peptides. Short amino acid chains that latch onto specific proteins. Vaccine work. Drug development stuff. The AI usually guesses based on data it’s seen before. But the quantum hybrid—mixing quantum qubits with traditional processors—pushed the model into uncharted territory. Especially where data was thin.
Is this a revolution? Not yet. Quantum computers are still tiny. Clunky. Expensive. But it’s a start. A proof of concept. Jenkins wasn’t buying it at first. He was a huge skeptic. Laughing now, but then, he thought applications were decades off.
The team focuses on proteins for immunotherapy. Cheap. Fast. Often funded by Novo Nordisk. Their headache? Bias in the data. Medical research loves Western genomes. Ignore the rest. Peptides designed there might fail elsewhere. In Africa. In Asia. The quantum approach generated more diversity. Even with scarce training data. That matters.
Still, don’t get excited yet.
Jonathan Funk, a DTU PhD student, says they couldn’t model a full-sized antibody. Not nearly complex enough. This is just one step. Finding a peptide isn’t curing the disease. It’s not a magic wand.
ORCA’s CEO, Richard Murray, agrees. Industry sees quantum as “hazy.” Distant. Why? Because clear, near-term uses are rare. This study? Novel. It shows commercial application now. They’re also talking to BP about chemistry and Toyota about car design efficiency.
Jenkins wants more. Bigger models. Larger proteins. Neglected diseases need this. No funding for snakebite antidotes usually. Until now? He’s trying to use quantum to design synthetic venom antidotes. A long shot. Maybe.
It’s messy. Real science often is. The needle might move. Slowly. But it might.
