Hi everyone,
I'm very interested in using the AMOEBA FF. I need to equilibrate NPT and NVT before production run. However I saw in OpenMM documentation that "The virial is not calculated for any force." Can you do NPT in OpenMM with AMOEBA?
Thanks!
AMOEBA FF for NPT, NVT
- Peter Eastman
- Posts: 2583
- Joined: Thu Aug 09, 2007 1:25 pm
Re: AMOEBA FF for NPT, NVT
Yes. Add a MonteCarloBarostat to your system. See http://docs.openmm.org/latest/userguide ... e-coupling.
- Daniel Konstantinovsky
- Posts: 77
- Joined: Tue Jun 11, 2019 12:21 pm
Re: AMOEBA FF for NPT, NVT
Thank you! Was worried barostat wasn't supported for AMOEBA.
One more q: John Chodera wrote in a old chat "Note that there can be issues when using finite-timestep MD without Metropolization together with a Monte Carlo barostat. The MC samples from the exact Boltzmann distribution, while the MD samples from a perturbed distribution in a timestep-dependent manner. Usually, this error is small, but with AMOEBA, it caused large problems."
Would using Langevin integrator address this?
One more q: John Chodera wrote in a old chat "Note that there can be issues when using finite-timestep MD without Metropolization together with a Monte Carlo barostat. The MC samples from the exact Boltzmann distribution, while the MD samples from a perturbed distribution in a timestep-dependent manner. Usually, this error is small, but with AMOEBA, it caused large problems."
Would using Langevin integrator address this?
- Daniel Konstantinovsky
- Posts: 77
- Joined: Tue Jun 11, 2019 12:21 pm
Re: AMOEBA FF for NPT, NVT
Actually never mind just saw the documentation re Langevin integrator