QuantaMind Study Reports DFT Level Accuracy for Reactive Molecular Simulations

MoleculeMind says in an announcement that research on its QuantaMind reactive atomistic modeling platform was published in Science Advances. The study reports Density Functional Theory level accuracy while running reactive molecular dynamics simulations lasting tens of nanoseconds for biomolecular systems containing 10,000 atoms.

QuantaMind simulates enzyme reactions including proton transfer, bond breaking, bond formation, and catalytic cycles. MoleculeMind says the platform has also been scaled to systems containing hundreds of thousands of atoms and can simulate one time step for a 100,000 atom reactive system in 0.25 seconds.

The platform uses a machine learning force field centered on transition states and trained with non equilibrium conformations and reaction data. It also incorporates classification embeddings for Density Functional Theory methods.

MoleculeMind describes applications in enzyme engineering and drug discovery. It says that in a project on pH sensitive antibodies, experimental testing found one candidate dissociated about 62 times faster at pH 6.0 than at pH 7.4. The paper is open access in Science Advances.