[gmx-users] Highest possible temperature in NPT?

[Berk Hess]

>From: pascal.baillod at epfl.ch
>Reply-To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>To: gmx-users at gromacs.org
>Subject: [gmx-users] Highest possible temperature in NPT?
>Date: Tue, 02 May 2006 16:39:57 +0200 (MEST)
>
>Dear community,
>
>I am trying to perform some wild REMD tests, and see the effect of very 
>high
>temperature on a solvated 100-residue protein.. ("hottest" replica 
>temperatures
>of 500-1000K). So far, the simulations would crash whenever using the NPT
>ensemble.. temperatures would remain correct but the volumes would grow too
>big. In NVT, I can apparently go as high in temperature as I wish. I 
>suppose I
>could get it to work for NPT as well, provided a massive increase of the 
>tau_p
>pressure time coupling constant. I found no info on that in the mailing 
>list
>and was wondering if anybody has experience on
>
>a) How much can I increase tau_p without getting too unphysical (under the
>assumption there is still anything physical at >500K!)? I read that the
>important thing is to conserve the compressibility/tau_p ratio.. should I
>change tau_p AND compressibility values for higher temperatures?

I assume your solvent is water and water is supposed to be
in the gas phase above 500 K. So I don't see the point in
increasing tau_p, as the stable configuration would be a protein
in a water gas.

Berk.