Minimizing memory as an objective for coarse-graining.

Area of Science:

• Computational Chemistry
• Molecular Dynamics
• Statistical Mechanics

Background:

• Coarse-graining reduces molecular complexity by integrating out degrees of freedom.
• Selecting appropriate coordinates is crucial but challenging.
• Failure to separate timescales leads to memory effects in reduced models.

Purpose of the Study:

• To introduce a heuristic measure of memory for comparing coordinate selections in coarse-graining.
• To provide an objective method for guiding the coarse-graining process.

Main Methods:

• Development of a novel heuristic measure of memory.
• Numerical exploration of the heuristic on three systems: a linear four-particle model, a nonlinear sixteen-particle model, and an atomic-resolution protein (actin monomer).

Main Results:

• The heuristic successfully identified physically intuitive coordinate selections across all tested systems.
• Results demonstrate the heuristic's utility in comparing different coarse-graining strategies.
• The heuristic reflects underlying molecular structure.

Conclusions:

• The memory heuristic serves as an objective tool for coarse-graining coordinate selection.
• It effectively codifies expert knowledge and highlights areas needing further attention in molecular models.