From: Ajasja Ljubetič (
Date: Wed Feb 16 2011 - 03:20:25 CST

Hi Sharlene,

unless you have a *huge* system I would recommend leaving some water padding
a the edge of the periodic cell to avoid artifacts due to PBC. I always
leave at least 5-10 A water padding.
If you would still like to do this, perhaps you could manually move the
protein to the neighboring cell and mark the waters fro removal using "water
and within 5 of protein". An example of removing unwanted waters in given on
page 29 of the membrane
There are 26 neigboring cells, but hopefully the protein is not wraped to
all of them.

So it would go something like this, which I typed of the top of my mind (a,
b ,c are the PBC vectors, you would have to repeat this for different
combinations of PBC vectors)

Best regards,

set prot [atomselect top protein]
#move the protein
$prot moveby $a

#get the ids of overlaping waters
set all [atomselect top all]
$all set beta 0

set badwater [atomselect top "water and within 1 of protein"]
set seglist [$badwater get segid]
set reslist [$badwater get resid]

mol delete all
package require psfgen
readpsf my.psf
coordpdb my.pdb

foreach segid $seglist resid $reslist {
delatom $segid $resid

writepdb my.pdb
writepsf my.psf

On Wed, Feb 16, 2011 at 07:03, Sharlene Denos <> wrote:

> Hello,
> I am trying to run a NAMD simulation using periodic boundary conditions
> (PBC). My initial configuration includes a water box that is slightly
> smaller than the total span of the proteins solvated. This should not be a
> problem for the simulation due to the use of PBC except that the water box
> does not have room for the proteins to wrap around on the other side. I
> created my water box using the Solvate plugin in VMD and I did not see an
> option for PBC. Does this exist? If not, what's the easiest way to figure
> out which waters need to be deleted from my system in order to make room for
> the protein to be wrapped around?
> Thank you!
> Sharlene