Accounting for hydrodynamic interactions in coarse-grained protein model
Moscow Institute of Physics and Technology, Moscow, Russia
1University of Massachusetts, Lowell, MA, USA
Computational (is silico) experiments are an important tool for exploring microscopic features of macroscopic processes in biomolecules, e.g. mechanochemical transitions in AFM experiments. But the computational costs associated even with coarse-grained Molecular Dynamics (MD) simulations are often unacceptably high because of large difference between typical timescales of in vitro and in silico experiments.
Self Organized Polymer model (SOP) is a native topology based coarse-grained force field for biomolecules. The SOP-GPU package employing Graphics Processing Units makes it possible to follow the dynamics of biomolecular system comprising up to 105 aminoacid residues in the experimentally-relevant subsecond timescale. Herein we feature the extended SOP model with hydrodynamic interactions fully implemented on GPU, in which solvent-induced many-body effects are accounted for using Rotne-Prager-Yamakawa tensor combined with the Truncated Expansion formalism.
We demonstrate the model's capability by performing the simulations of protein forced unfolding and indentation for specific examples of protein systems of different levels of complexity from a small protein (WW domain; N≈101 amino acids) to a large biological assemblies (virus particles CCMV and HK97; N≈104–105). We demonstrate that despite additional computational cost associated with accounting for hydrodynamic interactions, the total computation time is reduced due to accelerated kinetics, while majority of microscopic features are preserved.