Coarse-graining molecular dynamics: stochastic models with non-Gaussian force distributions

Area of Science:

• Computational Biology
• Biophysics
• Multiscale Modeling

Background:

• Cellular behavior modeling faces challenges due to scale differences between atomic and cellular processes.
• Multiscale methodologies use molecular dynamics (MD) and coarse-grained models to bridge these scales.
• The accuracy of coarse-grained models is crucial for reliable multiscale simulations.

Purpose of the Study:

• To present a new family of nonlinear stochastic coarse-grained (SCG) models.
• To address the limitations of linear models in capturing non-Gaussian force distributions from MD simulations.
• To develop SCG models that are easily parameterized using MD data.

Main Methods:

• Developed nonlinear stochastic differential equations for SCG models.
• Incorporated non-Gaussian force distributions observed in MD simulations.
• Derived solutions for the SCG model using gamma functions.

Main Results:

• The nonlinear SCG models accurately approximate detailed MD descriptions.
• Non-Gaussian force distributions, missed by linear models, are successfully incorporated.
• Parametrization of SCG models using MD simulations is not complicated by nonlinearities.

Conclusions:

• Nonlinear SCG models offer an effective approach for multiscale modeling in cellular biology.
• These models enhance the accuracy of coarse-grained simulations by including non-Gaussian dynamics.
• The developed models provide a computationally tractable method for bridging atomic and cellular scales.