DPG Phi
Verhandlungen
Verhandlungen
DPG

Berlin 2024 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

CPP: Fachverband Chemische Physik und Polymerphysik

CPP 23: Modeling and Simulation of Soft Matter III

CPP 23.5: Talk

Wednesday, March 20, 2024, 10:30–10:45, H 0107

Machine learning of an implicit solvent for dynamic Monte Carlo simulationsAnkush Checkervarty1, Jens-Uwe Sommer1,2, and •Marco Werner11Institut Theorie der Polymere, Leibniz-Institut für Polymerforschung Dresden e.V., Dresden, Germany — 2Institut für Theoretische Physik, Technische Universität Dresden, Germany

We discuss an implicit solvent model based on an artificial neural network (NN) for dynamic Monte Carlo simulations, where the dynamics is implemented only via local particle displacements (elementary motions). The training data was obtained from coarse grained simulations using the bond fluctuation model with explicit solvent [1] for single homopolymers under variation of solvent quality. The NN based implicit solvent model takes into account only the information of the local environment of monomers in order to predict a distribution of possible acceptance rates [2] of an attempted elementary monomer move in the given configuration. We show that NN-based implicit solvent simulations reproduce the coil-globule transition, as well as dynamic properties such as the bond vector autocorrelation in time and mean square displacements of chain monomers as seen in the explicit solvent model. Furthermore, the learned NN parameters were transferable to a system of multiple homopolymers. [1] C. Jentzsch, M. Werner and J.-U. Sommer, JCP 138, 094902 (2013). [2] A. Checkervarty, J.-U. Sommer and M. Werner, JCP 158, 124904 (2023).

Keywords: neural network; implicit solvent; monte carlo; coarse graining; machine learning

100% | Mobile Layout | Deutsche Version | Contact/Imprint/Privacy
DPG-Physik > DPG-Verhandlungen > 2024 > Berlin