Re: Vacuum simulations

From: Aron Broom (broomsday_at_gmail.com)
Date: Mon Oct 15 2012 - 15:14:28 CDT

NPT means constant: number of atoms, pressure and temperature. In a vacuum
there is essentially no pressure, so you would never use NPT for vacuum
simulations. Rather NVT would be what you probably want.

In terms of PBC, I suspect you could do that, but are you sure you want
to? Normally one does PBC such that the calculation is tractable insofar
as not having an obscene number of water molecules around the protein.
Unless you really believe the gas phase concentration of your protein is so
high that it warrants being treated in this periodic manner, I wouldn't use
that. Moreover, PBC should only be used when the net charge of your system
is zero, which is unlikely to be the case for you (since you have no
solvent and hence, no counter ions).

I'd suggest just using a langevin thermostat, with a small damping
coefficient (1/ps for instance). The one important modification you need
to make is electrostatic cutoff. Normally one uses some kind of
electrostatic cutoff, beyond which PME is used, but that requires PBC, so
you'll want to not have a cutoff at all and actually calculate all the
electrostatic interactions.
So you want to set cut to something larger than any distance between atoms
of your protein could be (e.g. 999), and then turn "switching off" and
"staticatomassignment yes" to save on doing pair-list calculations which
are now meaningless because of the cutoff.

If you have a moderately sized protein this will be a good solution. If
your protein is huge, the fact that you need to calculate all the pairwise
interactions because you now lack PME, can be devastating to performance.
In that case you'd have to come up with a clever way of neutralizing your
system so that you can use PBC and PME (e.g. dummy atoms), unless you are
fortunate enough that the species of interest has zero charge.

~Aron

On Mon, Oct 15, 2012 at 3:16 PM, mohan maruthi sena
<maruthi.sena_at_gmail.com>wrote:

> Hi all,
> I want to perform vacuum simulations for a protein using
> NAMD under NPT conditions. I am not sure whether I can use PBC /NPT
> conditions for vaccum simulations or not. I am sceptical about using PBC
> conditions. Can any one suggest whether , I can use NPT and PBC
> conditions for vaccum simulations.
>
>
> Thanking you,
> Mohan
>

-- 
Aron Broom M.Sc
PhD Student
Department of Chemistry
University of Waterloo

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:22:10 CST