Researchers at the Icahn School of Medicine at Mount Sinai have identified a previously hidden druggable site in a cancer-related protein that could open the door to developing a new generation of more precise cancer drugs. This discovery also reveals important limitations of today’s artificial intelligence tools for drug discovery.
The study was published online June 2. Journal of the American Chemical Society (10.1021/jacs.6c05178) focused on PKMYT1, a type of protein known as a kinase that helps control cell growth and division. Because this process can go awry in cancer, PKMYT1 has emerged as a promising target for new anticancer drugs.
Most experimental drugs designed to block kinases work by targeting a region called the ATP-binding site, a part of the protein that uses the cell’s energy supply to function. However, many kinases share nearly identical ATP binding sites, making it difficult for drugs to distinguish between their target of interest and other kinases, which can lead to undesirable side effects.
Using a combination of AI-based protein prediction tools and laboratory experiments, researchers discovered an entirely new “hidden” pocket in PKMYT1 to which the molecule can bind. This pocket was missed by today’s most advanced AI systems.
Our study demonstrates both the power and limitations of AI in drug discovery. Although the AI was highly accurate in predicting the shape of known proteins, it missed completely unexpected binding pockets that could only be discovered experimentally. That hidden site could ultimately provide a new way to design more selective anticancer drugs. ”
Dr. Avner Schlessinger, co-senior author and co-corresponding author, professor of pharmacology, director of the AI Small Molecule Drug Discovery Center, and associate director of the Mount Sinai Center for Therapeutic Discovery at the Icahn School of Medicine at Mount Sinai
The findings suggest that proteins such as PKMYT1 are much more flexible than previously recognized, constantly shifting between different forms rather than existing in a single, fixed form. The researchers said the study also found that even small chemical changes to a molecule can dramatically change how and where that molecule binds to proteins.
The research team used the AI system AlphaFold2 to predict possible structures of PKMYT1 and performed a virtual screen to identify molecules that might interact with it. They followed up with X-ray crystallography, biochemical tests, and cell studies to see how the molecules behaved in different experimental systems.
We then used additional AI tools such as AlphaFold3 and Voltz-2 and molecular dynamics simulations to test whether the current computational approach could predict the newly discovered binding modes.
“One of the most surprising findings was that a very small chemical modification switched the molecule from binding in this hidden pocket to binding in a more conventional manner,” says co-senior and co-corresponding author Michael Lazarus, Ph.D., associate professor of pharmacology and associate director of the Mount Sinai Center for Therapeutic Discovery at the Mount Sinai School of Medicine. “This shows that these proteins are incredibly dynamic and sensitive to subtle molecular changes, and also supports why experimental validation remains essential even in the age of AI.”
The researchers say this work could ultimately help scientists develop more selective drugs that avoid some of the toxicity and specificity challenges associated with traditional kinase inhibitors. This discovery could also help improve future AI systems by teaching them to better recognize hidden and dynamic protein states that are currently overlooked.
Although additional research is needed, this finding provides an important initial basis for developing future treatments targeting this newly discovered site. The compounds identified in the study represent a promising starting point for further optimization and testing in disease models.
Next, the team plans to develop more potent compounds that target the newly discovered site and investigate whether similar hidden pockets exist in other cancer-related kinases. They also hope to improve their computational methods in the future so that AI systems can more accurately predict the shape of these hard-to-detect proteins.
sauce:
Mount Sinai Health System
Reference magazines:
Herrington, New Brunswick; others. (2026). Allosteric inhibition of PKMYT1 induces a unique inactive ATP-binding site conformation. Journal of the American Chemical Society. DOI: 10.1021/jacs.6c05178. https://pubs.acs.org/doi/10.1021/jacs.6c05178
