Developing Machine-Learned Potentials for Coarse-Grained Molecular Simulations: Challenges and Pitfalls
This work addresses the problem of enabling larger-scale and longer-timescale molecular simulations for researchers in computational chemistry and materials science, but it is incremental as it builds on existing methods like SchNet.
The paper tackled the challenge of developing machine-learned potentials for coarse-grained molecular simulations by applying SchNet models to liquid benzene, exploring how model architecture and hyperparameters affect thermodynamic, dynamical, and structural properties, and reporting on encountered challenges and future directions.
Coarse graining (CG) enables the investigation of molecular properties for larger systems and at longer timescales than the ones attainable at the atomistic resolution. Machine learning techniques have been recently proposed to learn CG particle interactions, i.e. develop CG force fields. Graph representations of molecules and supervised training of a graph convolutional neural network architecture are used to learn the potential of mean force through a force matching scheme. In this work, the force acting on each CG particle is correlated to a learned representation of its local environment that goes under the name of SchNet, constructed via continuous filter convolutions. We explore the application of SchNet models to obtain a CG potential for liquid benzene, investigating the effect of model architecture and hyperparameters on the thermodynamic, dynamical, and structural properties of the simulated CG systems, reporting and discussing challenges encountered and future directions envisioned.