Merck and Mayo Clinic team up on AI-driven drug discovery

Two health industry institutions whose collaborations date back to the 1940s are teaming up on a project for a new era.

Merck and Mayo Clinic announced an agreement today that will allow the drug giant to tap the health system’s genomic and clinical data for AI and advanced analytics in drug discovery and development. The orgs claim the partnership marks Mayo’s first deal of this scale with a pharma company.

The partnership hinges on the Mayo Clinic Platform, which offers data from the US hospital and its international partners, like lab results, medical imaging, clinical notes, and molecular data.

“[Mayo has] a really unique wealth of de-identified clinical, molecular multimodal data sets, and these are not readily available in the healthcare landscape, at least in a really clean and highly curated way,” Greg Hersch, Merck’s SVP of enterprise strategy and venture, told Healthcare Brew.

Both parties say deals like these between health systems and pharma companies are critical to unlocking the potential of AI-driven drug discovery and development. Each org is also exploring future similar partnerships, Hersch and Maneesh Goyal, chief operating officer for the Mayo Clinic Platform, both told us.

“The whole platform concept was born out of our CEO looking at other industries and how they’ve embodied platform thinking, which is shared resources, collaborative models, modular thinking,” Goyal said. “Healthcare has been fairly against this, because there’s proprietary contracts, proprietary data sources, proprietary everything.”

The partnership will initially focus on developing therapies for three areas: inflammatory bowel disease, atopic dermatitis, and multiple sclerosis. Goyal said the strategy was to “look at the entire data set and identify where there’s opportunity, where there’s gaps in coverage that are aligned with Merck’s core areas.” They arrived on five areas that way then ultimately narrowed it to these final three, he said.

“What we did is we started engaging with [Mayo] and we looked at a lot of different scientific hypotheses,” Hersch said. “These were just the ones that lined up the best with our areas of research, and where Mayo felt that they could make a real foundational difference.”

Goyal said the Mayo Clinic Platform allows researchers to look beyond narrow data sets on specific patient cohorts and gain a bigger-picture understanding of disease progression that encompasses sources like deep analysis of clinical notes.

“All of the clinical data and omics data and waveform data all together will give us a view on disease progression unlike any time in history, and AI is just an accelerant toward that,” Goyal said. “We’re just at the precipice of looking at this data and saying, ‘What have we missed that’s been right in front of us?”

For Merck, the partnership is just one aspect of how it’s adopting AI. The company recently teamed up with Nvidia on a small-molecule drug discovery model called KERMT, and previously created a genomics-focused foundation model called TEDDY. Merck has also been integrating AI into areas like “regulatory responses, supply chain, forecasting, [and] commercial operations,” Hersch told us.

“We don’t go out a lot and talk about it, but we have a pretty end-to-end approach going on with technology and AI here at Merck, and we’re very proud of the progress we’ve made, but we understand that there’s a whole bunch more to do,” Hersch said.