Re: [AMBER] aMD and scaled MD simulation-NVT production runs

From: Vlad Cojocaru <>
Date: Mon, 22 Jun 2015 09:43:57 +0200

Well, first I would play with that in classical simulations before
running enhanced sampling stuff ... I would always advice to first run
a classical simulation stint in the desired ensemble before starting any
enhanced sampling simulations. And I would always start an enhanced
sampling simulations in the same ensemble as the preceding classical
simulation. You need the classical simulation to first understand the
behavior of your system before introducing perturbations. This also
ensures that you have a control simulation with which any further
enhanced sampling simulations can be compared with.

Normally, when you change from NVT to NPT there is not much that can go
wrong ... However, when you change from NPT to NVT, if the restarting
frame you use has a volume that is at either edge of the histogram of
the volume calculated from the previous NPT run, you might see a drift
in the average pressure which you want to avoid ....

What I normally do is plot the Volume histogram from the NPT run and
choose the volume that corresponds to the histogram peak. You can choose
a frame from the end of the equilibration which has a volume as close as
possible to the desired value or you could also take last frame from the
equilibration with the box coordinates modified to fit the desired
volume. Then I run an additional equilibration in the NVT ensemble with
the chosen volume and I monitor closely the pressure and energy terms
.... If I don't see drifts, I start a production run ...

But this is one way to do it, other people may have other (e.g. better)
recipes. I believe that important is to avoid drifts in the average
pressure under NVT conditions to ensure that no artificial tension in
introduced in your system.

Hope this helps

On 06/22/2015 08:24 AM, Neha Gandhi wrote:
> Dear List,
> I have seen papers on on aMD, Scaled MD and at times REMD where final
> system equilibration wis done in isothermal–isovolumetric (NVT) ensemble
> followed by 400 ps NPT run to ensure that the water box of simulated
> systems had reached the appropriate density. Then, production simulations
> were performed in the NVT ensemble.
> I would like to know more about ensemble switch to NVT from NPT during
> production? Can we start NVT run from with the last frame using the average
> density from NPT (langevian thermostat)? How do I take care of the volume,
> density and pressure while changing the ensemble?
> Thank you

Dr. Vlad Cojocaru
Computational Structural Biology Laboratory
Department of Cell and Developmental Biology
Max Planck Institute for Molecular Biomedicine
Röntgenstrasse 20, 48149 Münster, Germany
Tel: +49-251-70365-324; Fax: +49-251-70365-399
Email: vlad.cojocaru[at]
AMBER mailing list
Received on Mon Jun 22 2015 - 01:00:02 PDT
Custom Search