Answers are grounded in this paper and cite their sources.
We present a procedure for the automated mapping and parameterization of coarse-grained (CG) models for dissipative particle dynamics simulations. In our mapping methodology, a given molecule is represented as a graph, and its CG representation consists of a renormalized graph in which groups of vertices (representing heavy atoms) are contracted to single vertices (representing CG particles). Our mapping algorithm identifies functional groups in the molecule and leverages efficient graph-theory-based algorithms to find an optimal mapping for a given mapping scale. This mapping is subject to the constraint that certain functional groups remain intact within the resulting CG particles. To parameterize the corresponding CG force field, we apply an iterative Boltzmann inversion-like procedure to determine parameters for the bond, angle and nonbond potentials. This is accomplished by fitting valence geometries, and the integral of the pair distribution functions for the CG model to an all-atom reference simulation. We validate our approach by automatically calculating octanol-water partition coefficients for a dataset consisting of 454 simple organic compounds. Comparing these to experimental values results in a correlation coefficient of 0.92, indicating the robustness of our coarse-graining procedure in producing viable CG models. In addition, we demonstrate the utility of the procedure on a drug formulation involving a nano emulsion reproducing the size of the globules observed in experiment.
No takes yet. Share an insight, caveat, or question.
Grzetic et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: