The relative entropy is fundamental to adaptive resolution simulations

Area of Science:

• Computational physics and chemistry
• Soft matter simulation
• Multiscale modeling

Background:

• Adaptive resolution techniques enable efficient simulations of soft matter systems.
• These methods couple atomistic and coarse-grained (CG) force fields, with particles changing resolution dynamically.
• A hybrid transition region facilitates the coupling between different resolution domains.

Purpose of the Study:

• To investigate the utility of relative entropy in understanding adaptive resolution simulation methodologies.
• To demonstrate how minimizing relative entropy can improve the coupling between atomistic and CG regions.
• To establish a quantitative relationship between the hybrid region's width and coupling seamlessness.

Main Methods:

• Application of relative entropy as a metric for adaptive resolution simulations.
• Development of coarse-grained potentials derived from minimizing relative entropy.
• Analysis of the hybrid region's width and its impact on simulation coupling.

Main Results:

• Relative entropy proves effective for analyzing adaptive resolution simulation techniques.
• Coarse-grained potentials minimizing relative entropy lead to smoother transitions between simulation regions.
• A quantitative relationship was derived linking hybrid region width to coupling seamlessness.

Conclusions:

• Relative entropy is a valuable tool for both understanding and optimizing adaptive resolution simulations.
• Minimizing relative entropy enhances the seamlessness of particle description changes in multiscale simulations.
• The findings provide guidelines for optimally configuring adaptive resolution simulations for soft matter systems.