OpenMM Toolkit SandCat Speeds Up Prospective Hunting for Hidden Protein Pockets
- SandCat ports the SWISH-X enhanced-sampling approach for cryptic pocket detection, previously tied to GROMACS and PLUMED, to the OpenMM engine, and adds standardised protocols for targets with no known hidden sites.
- It reproduces earlier SWISH-X outcomes while running roughly twice as fast.
- On difficult targets with characterised cryptic sites, every known site ranked among the leading candidates, while a negative-control protein showed none.
- This is an unreviewed preprint; the authors suggest the speed gain could allow large-scale use on hard-to-drug proteins.
💡 Points Based on the abstract
Porting SWISH-X to OpenMM matters less for the speedup than for removing the GROMACS/PLUMED patching burden, so a project team might actually run it. But the validation is retrospective: known sites recovered, one negative control. Without experimental confirmation of a newly found pocket, ligandability remains unproven.
SandCat: OpenMM Framework for Prospective Cryptic Pocket Discovery ↗