AW: box dimension for unwrapped boxes?

From: Norman Geist (
Date: Thu Oct 04 2012 - 01:05:47 CDT

Hi Jaqueline,


I'd just like to add some comments.


1. I also don't wrap during simulation, it's more comfortable to wrap
it afterwards by ones individual needs.

2. The CHARMm format DCD files, generated by NAMD and VMD, always
contain the CHARMm extra block in front of each coordinate block. So, in
does contain the box dimension in XYZ for every frame. If you write your own
program, it's a good choice to use the DCD plugin from VMD or NAMD as it is
well tested and performing. As the file description you will find in the net
seems more to be a schematic view, it's hard to write a own dcd reader as
you will have to read the real format out off various dcd reading/writing
codes out there.

3. If you want to do the analyzes with VMD, after loading the DCD you
will be able to simply use the wrap command as VMD loads the CHARMm
extrablock and therefore knows the box dimensions for each frame. For the
analysis of density, you should stay with the usual wrap, means every atom
crosses the boundary, gets wrapped. For visualization purpose, it's better
to use the option -compound res to prevent from overlong bonds. Also of
course you will like to use -center com -centersel protein. For doing movies
it's also nice to prevent the system from shifting away by using the rmsd
trajectory tool in VMD to center all frames.


Good luck


Norman Geist.


Von: [] Im Auftrag
von Jacqueline Schmidt
Gesendet: Mittwoch, 3. Oktober 2012 17:33
An: Aron Broom
Cc: namd
Betreff: Re: namd-l: box dimension for unwrapped boxes?


Thanks Aron for your respond,

I actually find the way to do it by:


molinfo $molid get a (/b/c)





From: Aron Broom <>
To: Jacqueline Schmidt <>
Cc: namd <>
Sent: Wednesday, October 3, 2012 11:19 AM
Subject: Re: namd-l: box dimension for unwrapped boxes?

I think that either way is a just a post-processing problem. I think it's
generally better to wrap the coordinates and when making a movie or
something, just keep the peptide centered in each frame. This way the
coordinates for each atom can't become too large (although maybe this isn't
an issue with the binary DCD format).

Anyway, for your problem, I think the box dimensions for re-wrapping are
actually written inside the DCD file, and so there is a convenient method
for wrapping them with VMD. I could be wrong about that, but I suggest
searching through the mailing list, as this problem comes up somewhat


On Wed, Oct 3, 2012 at 11:14 AM, Jacqueline Schmidt
<> wrote:

Dear Namd users,

I have ran several simulations in NPT ensemble and I did not wrap the
molecules b/c I wanted to get smooth movement of the peptide (for making
nice movies from the simulation results). Now I want to get some density
distributions from some the system species but I have encounterd with the
problem than I should wrap back the coordinated in order to get the
distribution plots (I need exact box dimension for that purpose). However,
since the simulation has been run in NPT, the size of the box changes and
because I did not wrap the molecucles in the simulation the "measure minmax"
can not give me the box dimension anymore!!!


Also in the standard namd output, I can see that the volume is reported but
not the dimensions in x, y, z


can anyone please help? can I use these simulations to these density


looking forward to hearing something positive!!

thnkas, jackie

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

This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:22:39 CST