Coarse Grained MD Simulations of Soft Matter

In this chapter, we introduce the motif about building chemically specific coarse-grained (CG) effective potentials for molecular dynamics in soft matter science. We firstly present the basic ideas and statistical physics foundations of determining effective potentials of CG models, and commonly used methods including iterative Boltzmann inversion, force matching and machine learning based force matching methods. We also illustrate another framework of coarse-graining, originated from Mori-Zwanzig formalism including dissipative particle dynamics and its variants. In the end of this chapter, we introduce a Bayesian method that fits CG dynamics directly to the reference trajectory.