Merck announced a research collaboration with Isomorphic Labs, the artificial intelligence drug discovery company spun out of Google DeepMind, to design small-molecule drug candidates against an undisclosed cancer target. The partnership, revealed on November 19, positions Isomorphic alongside three of the top four drugmakers by revenue as strategic collaborators.
The collaboration arrives days after Isomorphic Labs raised $5 billion in funding, following a $2.1 billion commitment six months earlier. Merck will work with the AI-powered biotech to discover therapeutic candidates against an oncology target the partners describe as of “strategic importance,” though they have not disclosed the exact molecular focus.
Under the agreement, Isomorphic will receive an undisclosed upfront payment plus research, development, regulatory and commercial milestones and tiered royalties on any product that reaches the market. Isomorphic Labs holds an option to co-fund development in return for elevated royalties, giving the biotech a stake in the financial upside of successful candidates.
AI Platform Meets Pharmaceutical Expertise
Isomorphic Labs will apply its proprietary artificial intelligence drug design engine, IsoDDE, to generate computational predictions and molecular designs. The platform will perform computational work intended to guide the research process, producing molecular structures that the partners hope will accelerate candidate identification.
Merck will contribute proprietary datasets and handle all preclinical testing, clinical trials, manufacturing, and sales activities. The pharmaceutical giant brings substantial internal data capabilities and drug discovery experience to the table, creating what the partners describe as a complementary pairing of wet-lab and dry-lab functions.
“The key to successfully leveraging AI/ML for molecular invention lies in close coordination of wet-lab and dry-lab functions,” George Addona, head of discovery, preclinical development and translational medicine at MSD Research Laboratories, said in a statement.
“By combining MSD’s substantial internal data capabilities and discovery expertise with Isomorphic Labs’ powerful modeling and compute capabilities, we aim to identify new ways to design and create better candidates faster,” Addona added.
No Near-Term Treatment for Patients
For people living with cancer, the deal represents a signal about where drug research heads rather than a new treatment option. The work remains at the earliest stage of discovery, and no drug from this partnership has been tested in a laboratory animal or a person.
Addona told reporters that the target ranks as one Merck considers strategically important, describing it as a “target of high interest” to the company. Neither partner disclosed the size of the upfront payment, the milestone and royalty figures, or a timeline for human testing.
Sarah Skerratt, Isomorphic’s chief research officer, explained that partners able to carry candidates through development prove essential to the company’s mission. The division of labor allows Isomorphic to concentrate on computational drug discovery and molecular design while Merck oversees subsequent development and commercialization stages.
Isomorphic Assembles Pharma Powerhouse Network
Merck joins a growing list of pharmaceutical giants betting that Isomorphic’s modeling and compute capabilities can advance drug discovery. The London-based biotech disclosed a deal with Eli Lilly in 2024 and partnered with Johnson & Johnson, positioning itself at the center of an elite network.
In 2025, J&J and Lilly ranked as the first and third largest drugmakers by revenue, respectively, while Merck held the fourth position. Isomorphic also counts Novartis among its partners, meaning its publicly disclosed list of collaborators includes four of the top eight drugmakers globally.
The concentration of pharmaceutical firepower reflects investor confidence in the biotech’s approach. Investors placed Isomorphic in the spotlight early in the week, pumping billions into the company and validating its computational drug discovery platform.
AlphaFold Roots and Nobel Recognition
Isomorphic Labs draws its scientific credibility from its connection to AlphaFold, the AI program that predicts the three-dimensional shape of proteins from their building blocks. Demis Hassabis, Isomorphic’s founder and chief executive, shared the 2024 Nobel Prize in Chemistry for that groundbreaking work.
Understanding protein structures accelerates drug discovery by revealing how potential medicines might interact with their biological targets. The computational approach reduces the time and cost traditionally required to identify promising drug candidates through laboratory experimentation alone.
The technology allows researchers to evaluate millions of potential molecular structures rapidly, narrowing the field to the most promising candidates before expensive laboratory work begins. This computational screening represents one of the most tangible applications of artificial intelligence in pharmaceutical research.
Merck Expands AI and Research Partnerships
The Isomorphic agreement adds to a string of artificial intelligence and research deals by Merck this year. In February, the company announced a collaboration with Mayo Clinic to use de-identified clinical and genomic data to support drug discovery.
In September, Merck licensed an early-stage cancer drug candidate aimed at a KRAS mutation from a Shanghai company, broadening its pipeline through external partnerships. The KRAS target has emerged as a major focus in oncology research after proving difficult to drug for decades.
During a panel on Big Pharma deal-making at Fierce’s London Life Sciences Week event on Tuesday, Grace Han McMahon, who leads Pacific business development and licensing for Merck, identified the U.K. alongside Boston, San Francisco, and Asia as among the “key life sciences ecosystems” where the company bases its discovery scientists.
Where AI Enters the Discovery Process
The Merck-Isomorphic partnership illustrates where artificial intelligence enters the drug discovery process most effectively. Computational tools excel at the early identification and evaluation stages, where researchers must screen vast numbers of potential compounds to find candidates worth advancing.
Traditional drug discovery relies heavily on trial-and-error laboratory work, synthesizing and testing thousands of molecules before identifying promising leads. AI platforms compress this timeline by predicting which molecular structures will likely bind to disease targets effectively before scientists synthesize physical samples.
The approach combines experimental data with computational modeling, allowing researchers to learn from past successes and failures to guide future decisions. Merck’s contribution of proprietary datasets enables Isomorphic’s algorithms to train on real-world pharmaceutical research data rather than theoretical models alone.
The partnership structure reflects a broader industry trend where computational biotechs focus on discovery while established pharmaceutical companies handle the expensive and time-consuming clinical development process. This division of labor plays to each partner’s strengths and spreads financial risk across multiple parties.
