← Back to Home

Merck teams up with Isomorphic Labs to design small-molecule cancer drugs using experimental data and AI

The collaboration targets an undisclosed cancer target, with Isomorphic designing molecules and Merck taking over subsequent development. The agreement makes the drugmaker’s existing experimental data a starting point for design, but no validation results for candidate molecules have yet been provided.

By SURL BioNews

One key question in whether AI can help develop cancer drugs is how to turn data accumulated in the laboratory into new molecules that can be tested. Merck, also known as MSD, and Isomorphic Labs announced a research collaboration on November 19 to combine the drugmaker’s proprietary data with an AI drug design engine to identify small-molecule drug candidates against a cancer target.

According to the companies’ announcements, Isomorphic will use its IsoDDE engine, incorporating proprietary datasets selected by Merck for the collaboration, to generate computational predictions and deliver molecular designs. The target has not been disclosed, and the information currently available does not establish which cancer types the drugs might address or what treatment strategy is anticipated.

The project builds on Merck’s existing experimental work. According to STAT, George Addona, Merck’s head of drug discovery, preclinical development and translational medicine, said the company had previously conducted internal research on the target. This provides specific context for the source of the data, but the size of the datasets, what was measured, and how the data will be used in model design and evaluation have not been disclosed.

The division of responsibilities also defines AI’s role in this project: Isomorphic is responsible for predictions and molecular design, while Merck will undertake subsequent preclinical and clinical development, manufacturing and commercialization. In other words, computer-generated designs still require experimental testing to determine, step by step, whether the molecules have the required activity, selectivity and safety.

Under the commercial terms, Isomorphic will receive an upfront payment and will be eligible for research, development, regulatory and commercial milestone payments, as well as tiered royalties on net sales. The company may also choose to co-fund development in exchange for higher royalties. Dow Jones Newswires reported that the upfront payment amount was not disclosed, so the transaction cannot currently be used to assess the project’s valuation.

The announcements have not provided molecular structures, experimental activity data, animal study results or human trial results. They also have not presented comparative data demonstrating that the collaboration has shortened development timelines or improved the quality of drug candidates. What can be confirmed at this stage is that a new research project combining existing experimental data with AI design has begun. Whether it can produce cancer-fighting molecules worth advancing will depend on validation evidence disclosed subsequently.

References

  1. Merck
  2. Isomorphic Labs via PR Newswire
  3. STAT
  4. Dow Jones Newswires via MarketScreener