py-MCMD: Python Software for Performing Hybrid Monte Carlo/Molecular Dynamics Simulations with GOMC and NAMD

Area of Science:

• Computational chemistry and biophysics
• Molecular simulation and modeling
• Software development for scientific computing

Background:

• Molecular dynamics (MD) and Monte Carlo (MC) simulations are powerful tools for studying molecular systems.
• Integrating different simulation methods can overcome limitations and improve efficiency.
• Efficient workflows are crucial for managing complex simulation data and analysis.

Purpose of the Study:

• To introduce py-MCMD, an open-source Python software for hybrid MC/MD simulations.
• To validate the py-MCMD workflow by performing simulations for water and protein hydration.
• To demonstrate the computational efficiency and sampling advantages of the py-MCMD approach.

Main Methods:

• Developed py-MCMD as a workflow layer managing communication between NAMD and GOMC.
• Performed hybrid MC/MD simulations for SPC/E water in NPT and GC ensembles, and GEMC.
• Applied hybrid GCMC/MD simulations to hydrate a binding pocket in bovine pancreatic trypsin inhibitor.

Main Results:

• Hybrid MC/MD simulations showed close agreement with reference MC simulations, with 2-136x greater computational efficiency.
• Coupled-decoupled configurational-bias MC (CD-CBMC) algorithm in MC/MD simulations reached equilibrium 25 times faster for water in graphene pores.
• GCMC/MD simulations for protein hydration yielded water occupancies matching crystallographic data, with 5x better efficiency than MD.

Conclusions:

• py-MCMD provides a robust and efficient workflow for hybrid MC/MD simulations.
• The software enables significant gains in computational efficiency and sampling for various molecular systems.
• py-MCMD facilitates accurate modeling of complex phenomena, including protein hydration and confinement effects.