[gmx-users] on NVE and pressure

[Berk Hess]

>From: lorix <Loris.Moretti at pharm.unige.ch>
>Reply-To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>Subject: [gmx-users] on NVE and pressure
>Date: Mon, 16 Apr 2007 09:40:58 +0200
>
>Hi everybody,
>
>I am simulating a protein (approximately 5000 atoms) in water (35000 
>atoms). The protein is out of a thermodynamic equilibrium because I 
>modified a crystallographic structure of one of its conformation (which 
>showed stability with MD). Thus, my aim would be to let the system evolve 
>to a new equilibrium. I am expecting a conformation for the protein smaller 
>in size than the starting point.
>I had several simulations with Temperature and pressure coupling and now I 
>have decided to go for NVE (after a prior equilibration with TP controls 
>for 300 K and 1 bar) , the decision is mainly because I do not want to 
>affect the evolution of atom velocities during the simulation.
>With the NVE ensemble after 2 nsec I have an average  value for Temperature 
>about 285 K while the Pressure average value is around -230 bar. The 
>temperature is clearly fine. While the big negative value for pressure is 
>meaning that the system wants to contract  if I understood correctly ? How 
>does this pressure affect the system? Have I to consider the pressure just 
>like a tendency of the system either to contract or to expand?
>
>thanks for any suggestion
>loris...

You have equilibrated your modified conformation at 300 K and 1 bar?
Then it does not make sense to use NVE.
It only makes sense to use NVE when you have some ultra fast process
that somehow puts energy into your system and you have a way
to describe this energy transfer process reasonably accurately.

You can not judge if your ensemble is ok just by looking at the temperature.
In NVE you should make sure the total energy is reasonably conserved.

Berk.

_________________________________________________________________
Hotmail en Messenger on the move http://www.msn.nl/services