Re: NPAT - NPgT ensemble

From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Sat May 02 2009 - 20:18:46 CDT

Hi Bjoern,

Bjoern Olausson wrote:
>
> Now I got three questions:
>
> 1) In my CHARMM setup for NPAT (pmxx 0. pmyy 0. pmzz 2295) pmzz is coupling the
> box length (z) to the according component of the pressure tensor. Is the keyword
> "UseConstantArea" working in the same way? As much as I could see, there is
> no option to couple x or y to the according pressure tensor, so I guess NAMD
> expects a membrane to be in the X/Y-Plain having the Lipids aligned along the
> Z-Axis (Actually the standard setup).
>

This is correct.

> 2) Is my namd config for NPAT/NPgT right?
>

Looks reasonable to me. I don't speak charmm fluently; you might want to
consider useConstantRatio if you've got a flag to keep fluctuations in
the x and y directions isotropic.

> 3) Is it more efficient to use "PMEGridSize{X,Y,Z}" (having the factors 2,3 and
> 5 in mind) then "PMEGridSpacing"? Is "PMEGridSpacing" automatically calculating
> the GridSize with respect to the suggested factors?
>

At least in the current version of the code it looks like the
automatically calculated sizes follow the small prime factor rule, but
it will always give you more control to do it by hand.

Best,
Peter

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:50:52 CST