Replica-exchange simulation algorithms for OpenMM.
This package provides a general facility for running replica-exchange simulations, as well as derived classes for special cases such as parallel tempering (in which the states differ only in temperature) and Hamiltonian exchange (in which the state differ only by potential function).
Simulations utilize a generic Markov chain Monte Carlo (MCMC) framework that makes it easy to mix Monte Carlo and molecular dynamics.
Provided classes include:
ReplicaExchange
- Base class for general replica-exchange simulations among specified ThermodynamicState objectsParallelTempering
- Convenience subclass of ReplicaExchange for parallel tempering simulations (one System object, many temperatures/pressures)HamiltonianExchange
- Convenience subclass of ReplicaExchange for Hamiltonian exchange simulations (many System objects, same temperature/pressure)
Use of this module requires the following:
- Python 2.7 or later:
- OpenMM with Python wrappers:
- NetCDF (compiled with netcdf4 support) and HDF5 (on which NetCDF4 depends):
- netcdf4-python (a Python interface for netcdf4)
- numpy and scipy
- mpi4py (if MPI support is desired)
- http://mpi4py.scipy.org/
- Note that mpi4py must be compiled against the appropriate installed MPI implementation.
- mdtraj (optional)
- pandas
- OpenMMTools (https://github.com/choderalab/openmmtools)
- John D. Chodera [email protected]
- Kyle A. Beauchamp [email protected]
All code in this repository is released under the GNU Lesser General Public License. See LICENSE for details. Written by the authors at the University of California Berkeley and the Memorial Sloan-Kettering Cancer Center.