Re: [AMBER] hexagonal unit cell for membrane simulation with NAMD

From: Jason Swails <>
Date: Tue, 12 Nov 2013 17:06:30 -0500

On Tue, Nov 12, 2013 at 4:11 PM, Thomas Evangelidis <>wrote:

> One more detail. When I use in NAMD :
> wrapWater on
> wrapAll on
> wrapNearest off
> I get the hexagonal prism I showed, but when I set:

I presume you mean "wrapNearest on" here, right?

> wrapWater on
> wrapAll on
> wrapNearest off
> I get exactly the triclinic box you showed.

I had a conversation with Pawel about this and it's because these two
representations are equivalent. A hexagonal prism replicated periodically
can be represented as a simple triclinic cell with certain angles between
the unit cells (90, 90, and 120 degrees, for instance). Imagine cutting
off corners and attaching them to the opposite side in such a way that
every particle is represented with the periodic image closest to a
particular point in space. The imaging routines are generalized for
triclinic cells, so by default they will give you the triclinic cell Pawel
showed. If you instead take the image that is closest to a particular
point in space, then you will (necessarily) get the most spherically
symmetrical representation of that triclinic cell, which in your case is a
hexagonal prism.

This is the difference between using and not using the "familiar" keyword
in the (cp)ptraj image commands, I believe.


Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Tue Nov 12 2013 - 14:30:03 PST
Custom Search