QUELO 2.3.2 introduces hydrogen mass repartitioning (HMR) to QUELO-QM. QM potential energy surfaces are more complex than the corresponding classical examples therefore, HMR is much more challenging than that of a classical force field. With the successful implementation of QM-HMR we are able to cut time to solution in half while maintaining quantum-level accuracy by enabling larger simulation time steps.
By supplementing additional QM data to the training data set for dihedrals, we have enhanced our AI-FF model to deliver better accuracy for drug lead optimization. For example, molecules with biaryl structures, which are in many drug modalities, can now be reliably simulated using QUELO-AI. Leveraging AI has given us the ability to improve QUELO in ways no other platform can compete with.
Recent enhancements to the UI for Covalent FEP now allows users to employ adaptive QM water and uses an automated QM region selection that maps some of the protein residues in addition to the ligand. Furthermore, QUELO is now compatible with NVIDIA Blackwell GPUs. Making sure that our customers have exactly what they need to advance their pipelines is critical to us and we welcome feedback on ways to improve QUELO for you!
