TL;DR
A £20m hub is going up at the Cambridge Biomedical Campus to speed up drug development using AI alongside laboratory models grown from human tissue. The Medical Research Council is paying for it, the site sits within the university’s stem cell institute, and the local NHS trust is a working partner.
What the money buys
The premise is that most candidate drugs fail because the models used to test them do not behave like people. The hub will build models from human tissues and diseases — including organoids, the lab-grown mini-organs that mimic how real tissue responds — and apply AI to what those systems produce. The trust’s stated aim is sharper prediction and study of human disease, with treatment development happening before anything reaches a patient. Leaning less on animal research is a secondary goal.
Professor Matthias Zilbauer, who co-leads the hub and works as an honorary consultant at the trust, says models built around the characteristics of individual patients allow potential treatments to be assessed before the clinic. He expects the approach to make drug development quicker and cheaper, and personalised therapies more achievable.
Professor Bertie Göttgens, who directs the stem cell institute and shares the hub leadership, frames the value as convening power — pulling industry, research institutes and hospitals into one programme. Partners include Royal Papworth Hospital, the Wellcome Sanger Institute and the MRC Laboratory of Molecular Biology.
Why this one is different
Most UK AI funding announcements of the past year have bought compute or capacity: growth zones, sovereign compute allocations, university partnerships with chipmakers. This one buys biology. The constraint in drug discovery has never been a shortage of processing power to search chemical space — it is that the resulting predictions are validated against systems which poorly represent human physiology. Pairing model output with tissue that behaves like a patient’s own attacks the bottleneck that actually bites.
Looking forward
Cambridge already holds the densest concentration of UK life sciences infrastructure, so the hub reinforces an existing cluster rather than seeding a new one. The measure worth watching is not publication volume but whether candidates screened this way survive later-stage trials at a better rate than the current dismal average. That evidence will take years, and £20m is modest next to what pharmaceutical companies spend annually on failures.