Lundbeck turns to Cradle's AI for antibody discovery in brain disorders

Lundbeck, the Danish pharmaceutical company focused exclusively on brain health, will use Cradle's AI-powered platform to discover and optimise antibodies for central nervous system diseases, the companies announced in June. The partnership marks a shift for one of Europe's largest biotech firms from exploring AI as a research tool to embedding it directly into lead programs.

Cradle, the Amsterdam-based startup founded in 2021, offers a generative AI platform for designing proteins and optimising candidates. The tool operates through a feedback loop between computational design and laboratory testing. As scientists run experiments, their results feed back into the model, refining predictions for each subsequent round of design. Lundbeck will initially deploy the platform across two antibody programs targeting CNS diseases, aiming to reduce the number of expensive wet-lab cycles required to reach promising candidates. For Lundbeck's protein engineers and computational biologists, the approach means faster iteration: instead of months of testing to optimise an antibody, the model guides which variations are most likely to work.

The partnership reflects Lundbeck's broader push to embed AI across its research operations. Tarek Samad, the company's senior vice-president for global research and patents, said that Cradle's system "provides our scientists with powerful AI that fits seamlessly into our processes, strengthening our ability to discover and optimise therapeutics." Lundbeck has also formed a separate partnership with Denmark's Gefion AI supercomputer to accelerate structural modelling and broaden its search space for novel compounds. Together, these partnerships connect AI across Lundbeck's discovery pipeline, from initial target identification through optimisation of lead candidates.

Cradle has moved with speed into major pharma partnerships since raising $73 million in Series B funding five months earlier. The round, led by IVP and including Index Ventures and Kindred Capital, brought the company's total capital raised to over $100 million. Since closing that round, Cradle has announced deployments across more than fifty research programmes at eight of the world's twenty-five largest pharmaceutical companies. The company claims its platform can accelerate development cycles by two to twelve times, though real-world results depend on the specific protein target and assay constraints.

For Stef van Grieken, Cradle's chief executive and co-founder, the Lundbeck deal fits a pattern emerging across pharma. "The Cradle platform has been shown in internal and partner programs to accelerate design cycles significantly," he said. "By closing the loop between computation and the lab, we can help Lundbeck reach high-quality antibody candidates faster and with fewer iterations." The company, which maintains offices in Zurich and the United States alongside its Amsterdam base, draws its team of machine learning researchers and biologists from Google, Novartis, Meta and other technology and life-science firms.

Lundbeck's integration of Cradle into production workflows signals a broader shift in drug discovery, where computational tools have moved beyond research projects into core discovery processes. With half the world's population affected by brain disorders, Lundbeck said, accelerating the path from target identification to clinical-stage candidate matters for patients waiting for new treatment options. The Lundbeck partnership will test whether AI systems trained on protein data can genuinely compress years of optimisation cycles into months, a claim central to Cradle's pitch but one that still requires clinical validation.