| About this Abstract |
| Meeting |
MS&T24: Materials Science & Technology
|
| Symposium
|
Frontiers of Machine Learning on Materials Discovery
|
| Presentation Title |
Bayesian optimization of CG topologies: Applications to common polymers |
| Author(s) |
Pranoy Ray, Adam P. Generale, Nikhith Vankireddy, Yuichiro Asoma, Masataka Nakauchi, Haein Lee, Katsuhisa Yoshida, Yoshishige Okuno, Surya R. Kalidindi |
| On-Site Speaker (Planned) |
Pranoy Ray |
| Abstract Scope |
The realm of atomic systems constitutes the fundamental building blocks of matter and presents exciting potential for scientific and technological breakthroughs, and Molecular Dynamics (MD) as a physics-based simulation toolset, has set high standards for the accurate estimation of the physical and chemical properties of large molecular systems across varied pressure or temperature ensembles. Given, the high computational costs encountered in running classical MD simulations, the advent of Coarse-Grained Molecular Dynamics (CGMD) has helped advance molecular discovery at substantially lower costs. However, the lower-dimensional embedding of the coarse-grained molecular topologies by lumping multi-body effects into specialized potentials is a definite tradeoff between efficiency and accuracy. This work aims to explore the utilization of Bayesian Optimization methodologies towards rapidly refining the general-purpose CG-Martini3 topologies for domain-specific applications. The CGMD computations run with these optimized CG topologies aim to mimic the accuracy of classical MD while being faster by a few magnitudes. |