Recently, the Chung group at Southern University of Science and Technology (SUSTech) has combined efficient machine learning potentials (MLPs) with multi-scale quantum refinement methods to enhance computational efficiency and reliability. Their results are published in Nature Communications. The CC-quality AIQM1 method in MLatom@XACS software was …

Nat. Commun.:Accelerating reliable multiscale quantum refinement of protein-drug systems enabled by AIQM1 Read more »

Quasi-classical molecular dynamics (also known as quasi-classical trajectories (QCT)) accounts for the zero-point energy (ZPE) in contrast to classical dynamics and is very popular in studying chemical reactions (see, e.g., works by Houk et al. in PNAS 2012, 109, 12860 and J. Am. …

Quasi-classical trajectories to study reaction mechanisms like in PNAS and JACS papers! Read more »

We have held online broadcast on April 24, at 15:30 Beijing time/9:30 am CET on the XACS Youtube channel at https://www.youtube.com/watch?v=TOVmwgId-eA. In the broadcast, we have demonstrated how MLatom@XACS can be used for accelerating expensive quantum chemical simulations via efficient building …

View online broadcast: Active learning for building your data and machine learning potentials Read more »

A machine learning potential with low error in the potential energies does not guarantee good performance for the simulations. One of the reasons is that it is hard to train machine learning potentials with balanced descriptions of different PES regions, …

JPCL | Tell Machine Learning Potentials What They Are Needed For: Simulation-Oriented Training Read more »