GENESIS CGDYN: large-scale coarse-grained MD simulation with dynamic load balancing for heterogeneous biomolecular...

Area of Science:

• Biophysics
• Computational Biology
• Soft Matter Physics

Background:

• Coarse-grained molecular dynamics (CG MD) simulations are crucial for studying slow biological processes involving large biomolecular assemblies.
• Non-uniform distribution of biomolecules in large systems poses computational challenges for conventional simulation methods.

Purpose of the Study:

• To develop an efficient simulation technique for large, heterogeneous biomolecular systems.
• To investigate the dynamics and mesoscopic behavior of intrinsically disordered protein (IDP) droplets.

Main Methods:

• Developed a hierarchical domain decomposition scheme with dynamic load balancing for CG MD simulations.
• Applied the method to simulate the fusion and dynamics of multiple IDP droplets.
• Implemented the scheme in the CGDYN module of the GENESIS software package.

Main Results:

• The new method maintains computational efficiency even with dynamic changes in particle distribution.
• Observed a correlation between droplet shape changes and IDP chain mixing during droplet fusion.
• Directly simulated Ostwald ripening, where small droplets dissolve into larger ones, at mesoscopic scales.

Conclusions:

• The developed simulation strategy enhances the efficiency of large-scale biomolecular simulations.
• Provides a powerful tool for investigating mesoscopic phenomena like IDP droplet dynamics and phase separation.
• Enables simulations of biological systems at scales previously inaccessible to residue-level CG models.