Skip to the stories
Microsoft 365 & SharePoint

Isomorphic Labs and MSD Pair AI Drug Design With Oncology Data

Isomorphic Labs and MSD will combine AI drug design with MSD's data to find small-molecule cancer candidates, with payments tied to milestones.

Isomorphic Labs and MSD Pair AI Drug Design With Oncology Data

Isomorphic Labs, the Alphabet-owned drug discovery company, has signed a research collaboration with MSD aimed at finding new cancer drug candidates. It's a notable pairing: one side brings a generative and predictive design engine, the other brings large internal datasets and a long record of taking medicines through development. For anyone following AI in life sciences, the deal is a useful test of whether computational drug design can hand off cleanly to the wet lab and the clinic.

What the collaboration covers

The work centres on oncology drug therapies built around small molecules. The partners describe the goal as discovering therapeutic small molecule candidates against an oncology target of strategic importance. In plain terms, they will work together on a single prioritised target and try to find molecules that could become cancer treatments. The announcement doesn't name the target, so there's no public detail on the biology or the chemical series involved.

Top Google minds quit to build AI that accelerates research

How the division of labour works

Isomorphic Labs will handle the computational side. The company describes its approach as in silico prediction, meaning experiments run on computers rather than in glassware, used to generate candidate drug molecule designs. Its design engine, IsoDDE, will be fed MSD's datasets. That's the part I'd watch most closely. Pharma training data is often messy, siloed and tied to assay conditions that differ from one lab to the next, so how well the model absorbs MSD's data will largely determine whether the designs improve on what the company already has.

European Commission adds €500m to research and innovation funding

MSD takes on everything after the design stage: pre-clinical and clinical development, manufacturing, and bringing any resulting medicine to market. That split makes sense. Isomorphic Labs is building the design capability, and MSD brings the scale and regulatory experience needed to turn a molecule into a product. It also means the two companies will need to hand work back and forth across a fairly long pipeline, which is where collaborations like this often run into friction.

Dublin-headquartered Disseqt AI raises $8m in funding

The money and the milestones

Under the partnership agreement, Isomorphic Labs will receive an upfront payment from MSD. It will also be eligible for further payments tied to specified research, development, regulatory and commercial milestones. The source doesn't disclose the size of the upfront payment or the milestone amounts, so it's impossible to judge the deal's value from the announcement alone.

SomaTech takes top prize at All-Island Circular Venture Awards 2026

Milestone-based structures are common in drug discovery partnerships. They shift risk toward the party developing the molecule and reward the design partner only if candidates progress. For Isomorphic Labs, that means revenue depends on how far its designs travel, not just on whether the collaboration is signed.

Irish-founded End Close raises $3.5m in seed funding round

The funding behind Isomorphic Labs

The MSD announcement landed in the same week Isomorphic Labs said it had raised $5bn in Series C funding. The round was co-led by Greenoaks and Sequoia Capital, with Andreessen Horowitz and several other new investors, plus a significant investment from Alphabet. Existing investors Thrive Capital, MGX, GV, CapitalG, Temasek and the UK Sovereign AI Fund also took part.

Irish co-founded Pretxt raises $5m seed to supervise AI agents

That's a large amount of capital for a drug discovery company, and it puts the MSD deal in context. A round of this size suggests the company is planning long, expensive programmes rather than a quick proof of concept. Partnerships with established drugmakers can also help validate the technology, since a company like MSD has to be convinced the designs are worth testing in real patients.

What the two companies are saying

Sarah Skerratt, chief research officer at Isomorphic Labs, said: "Through this collaboration with MSD, we will apply our world-leading drug design engine to unlock significant discovery progress." She added that the work highlights the companies' shared commitment to innovation for the benefit of patient populations with unmet needs.

George Addona, senior vice-president and head of discovery, preclinical development and translational medicine at MSD Research Laboratories, put the emphasis on process. He said the key to using AI and machine learning for molecular invention lies in close coordination between wet-lab and dry-lab functions. MSD, he said, aims to combine its internal data capabilities and discovery expertise with Isomorphic Labs' modelling and compute capabilities to design and create better candidates faster.

Addona's point is the one I'd take most seriously. Plenty of AI drug discovery claims stall at the handoff between predicted molecules and synthesised, tested compounds. Whether this partnership delivers will depend on how quickly the computational predictions are checked against lab results and fed back into the design loop.

What to watch next

The announcement sets out the structure of the collaboration but not its timeline. There are no dates for candidate selection, pre-clinical work or clinical entry, and no stated success criteria. Those details will show whether the model is delivering faster or better candidates than conventional discovery, which is the claim both companies are implicitly making.

For readers outside pharma, the practical lesson is about data. The partnership depends on combining a design engine with a large company's proprietary datasets, and that kind of integration tends to be the hard part. Teams building AI pipelines on sensitive internal data should expect the same issues around data preparation, access control and feedback loops, whatever the sector. The next concrete signals will be any named programme milestones and whether the two companies report progress against them.

Blue and white capsule drugs against a yellow background.

Image: © Fahroni/Stock.adobe.com

Source: Laura Varley · www.siliconrepublic.com

Same flag