Deep Coarse-grained Molecular Modeling

Julija Zavadlav

Molecular modeling has become a cornerstone of many disciplines, including material science. However, the quality of predictions critically depends on the employed model that defines particle interactions. A class of models with tremendous success in recent years are neural network (NN) potentials due to their flexibility and capacity to learn many-body interactions. In this talk, I will present the current state-of-the-art in deep coarse-grained molecular modeling [1,2,3]. I will discuss the ongoing challenge of sufficiently large and broad training datasets and our approaches to alleviate this issue, including novel training objectives, combining different data sources, Bayesian uncertainty quantification, and active learning. I will showcase the effectiveness of these approaches for various test case biophysical systems.

1. Thaler, S. and Zavadlav, J., Nat. Commun. 2021, 12, 6884.
2. Thaler, S. and Stupp, M. and Zavadlav, J., J. Chem. Phys. 2022, 157, 244103.
3. Thaler, S. and Doehner, G. and Zavadlav, J., JCTC 2023, 19, 4520–4532.