[gmx-users] Use NVT to mimic NVE

[Michael Shirts]

Following up on fluctuations and ensembles:

The following paper discusses tests of whether an ensemble has the correct
distribution of energies, including fluctuations of all scale:

"Simple quantitative tests to validate sampling from thermodynamic
ensembles"
http://pubs.acs.org/doi/abs/10.1021/ct300688p

Turns out that v-rescale behaves very well, at least for fluctuations of
the total energy.

In the limit of large scale, for NVE and NVT (if carried out at the
temperature where <E>_NVT = E_NVE, then all system averages (including
pressure, etc) will be equal.  But the fluctuations (which include 2nd
derivative properties like heat capacity, etc). will not be equal.  So it
depends on what you are interested in.

If you are interested in thermodynamics, just go ahead and run the ensemble
you are interested in.  If you are interested in dynamics, you can check
out this paper which partly answers the questions of what the artifacts are
created by thermostats.  Again, vrescale is quite good --  Langevin or
anything with stochastic noise is not so good.

"Effects of Temperature Control Algorithms on Transport Properties and
Kinetics in Molecular Dynamics Simulations"
http://pubs.acs.org/doi/abs/10.1021/ct400109a


Best,
~~~~~~~~~~~~
Michael Shirts
Associate Professor
Department of Chemical Engineering
University of Virginia
michael.shirts at virginia.edu

(434)-243-1821
On Fri, Oct 10, 2014 at 7:40 PM, Antonio Baptista <baptista at itqb.unl.pt>
wrote:

> On Fri, 10 Oct 2014, Mark Abraham wrote:
>
>  On Fri, Oct 10, 2014 at 7:41 PM, Johnny Lu <johnny.lu128 at gmail.com>
>> wrote:
>>
>>  Hi.
>>>
>>> Is it a good idea to mimic NVE by a NVT simulation with a large
>>> temperature
>>> coupling time constant, to reduce the effect of the thermostat ?
>>>
>>>
>> Not if your observable is the total energy. But for most interesting
>> observables, this is more than fine.
>>
>>
>>  If I use V-rescale thermostat, what artifacts will the simulation get if
>>> I
>>> use a large coupling time (say, 500 ps) and single precision gromacs ?
>>>
>>>
>> Probably nobody knows :-) Thermodynamic ensemble differences disappear as
>> the system becomes large. But quantifying that from converged ensembles
>> for
>> a given observable is work.
>>
>> Mark
>>
>
> Note also that, unlike averages, fluctuations are ensemble-dependent.
> Although v-rescaling (as most other thermostats) does indeed sample from
> the NVE ensemble in the mathematical limit of infinite coupling constant,
> this does not imply that, in practice, fluctuations converge to those in
> the NVE ensemble for large but finite coupling constants.
>
> As you can check in the v-rescaling paper, fluctuations were found to be
> very robust to changes of the coupling parameter (at least for the kinetic
> and potential energies). This is generally a good thing, but maybe not what
> you want to run a "fake" NVE simulation as you described. So, if you are
> interested in the fluctuations of some properties, you will have to check
> what is in practice their asymptotic behavior.
>
> Best,
> Antonio
>
>
>
>
>>
>>  Thank you.
>>> --
>>> Gromacs Users mailing list
>>>
>>> * Please search the archive at
>>> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
>>> posting!
>>>
>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>>
>>> * For (un)subscribe requests visit
>>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>>> send a mail to gmx-users-request at gromacs.org.
>>>
>>>  --
>> Gromacs Users mailing list
>>
>> * Please search the archive at http://www.gromacs.org/
>> Support/Mailing_Lists/GMX-Users_List before posting!
>>
>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>
>> * For (un)subscribe requests visit
>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>> send a mail to gmx-users-request at gromacs.org.
>>
>>  --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/
> Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>