摘要
Computational investigations of biological and soft-matter systems governed by strongly anisotropic interactions typically require resource-demanding methods such as atomistic simulations.However,these techniques frequently prove to be prohibitively expensive for accessing the long-time and largelength scales inherent to such systems.Conversely,coarse-grained models offer a computationally efficient alternative.Nonetheless,models of this type have seldom been developed to accurately represent anisotropic or directional interactions.In this work,we introduce a straightforward bottomup,data-driven approach for constructing single-site coarse-grained potentials suitable for particles with arbitrary shapes and highly directional interactions.Our method for constructing these coarsegrained potentials relies on particle-centered descriptors of local structure that effectively encode dependencies on rotational degrees of freedom in the interactions.By using these descriptors as regressors in a linear model and employing a simple feature selection scheme,weconstruct single-site coarse-grained potentials for particles with anisotropic interactions,including surface-patterned particles and colloidal superballs in the presence of non-adsorbing polymers.We validate the efficacy of our models by accurately capturing the intricacies of the potential-energy surfaces from the underlying fine-grained models.Additionally,we demonstrate that this simple approach can accurately represent the contact function(shape)of non-spherical particles,which may be leveraged to construct continuous potentials suitable for large-scale simulations.
基金
funding from The Netherlands Organization for Scientific Research(NWO)for the ENW PPS Fund 2018-Technology Area Soft Advanced Materials ENPPS.TA.018.002 and for the OCENW.KLEIN.432,respectively
M.D.acknowledges funding from the European Research Council(ERC)under the European Union’s Horizon 2020 research and innovation programme(Grant agreement No.ERC-2019-ADG 884902=SoftML).
