Towards a priori uncertainty quantification in coarse-grained molecular dynamics: Generalized multipole potentials.

Area of Science:

• Computational Materials Science
• Molecular Dynamics Simulations

Background:

• Coarse-graining (CG) methods in computational materials science often lack upfront uncertainty quantification (UQ) tools.
• This absence hinders the adoption of CG techniques due to unknown model accuracy and calibration costs.

Purpose of the Study:

• To develop an analytical method for coarse-graining rigid-body systems.
• To provide a priori UQ for reduced-order models in CG molecular dynamics (MD).
• To establish a mathematical foundation for assessing CG force field quality without simulations.

Main Methods:

• Developed an analytical coarse-graining approach for rigid-body systems.
• Derived corresponding intermolecular potentials with controllable accuracy relative to atomistic models.
• Validated the method using simulated trajectories.

Main Results:

• The analytical method provides a mathematical basis for evaluating CG force field quality a priori.
• It allows understanding atomistic systems as limits of reduced-order models.
• Simulations confirmed the approach's validity at the trajectory level.

Conclusions:

• The presented method offers a way to assess CG model accuracy before extensive simulations.
• It addresses the need for UQ in coarse-grained molecular dynamics.
• Further work is needed for coarse-graining fully non-rigid systems.