From: Nicolas Sapay (
Date: Mon Mar 06 2006 - 10:52:23 CST

Dear Gonzalo,
As I just answer in an another thread, there is this possibility :

it should be possible to do this taking into account the PBC in your
selection command. Example: You want to select water within 5 A of a
- 1 : Make a selection using the "within" command:
set sel1 [atomselect top "water within 5 of resid 1"]
- 2 : If the residue 1 is at less than 5 A of a PBC (e.g. the PBC
related ot the x axis), translate the center of the first selection and
make a second selection:
# Measure the center of the first selection
lassign [measure center $sel1] a b c
# Translate the x coordinates taking in to account the length of your
set a [expr - (abs($a)/$a) * abs($length - $a)]
# Make a second selection considering that the cartesian equation of a
sphere is (x-a)² + (y-b)² + (z-c)² = r²
set sel2 [atomselect top "water and

Now, you just have to sum the number of atoms in the two selections :
puts "natom = [expr [$sel1 num]+[$sel2 num]]"

I'm not sure it will be easy to take into account the 3 dimensions with
this method.... However, I hope it help you!



Le lun 06/03/2006 à 16:23, a écrit :

    Maybe, too obvious but... about a bigger water box in the simulation? So the peptide won't move
    to the border walls of the box, or is the peptide too big?
    Another suggestion would be to fix one atom (as close as possible from the
    center of mass) in the simulation in order to null the translation effect
    (without affecting the rotation of the protein) to the walls.
    Date: Mon, 6 Mar 2006 01:06:50 +0100
    From: "L. Michel Espinoza-Fonseca" <>
    Subject: vmd-l: pbc and vmd
    Dear all,
    I've been trying to analyze the contacts between water and a given
    peptide from my trajectories. I used a nice script provided by one of
    the VMD users. Everything went well, except from the fact that VMD is
    not able to count the water molecules around my peptide from the image
    of my periodic box. I've read some issues related to the pbc and VMD
    (computation of hbonds, for instance), and according to what I read,
    there is not such handling of pbc in VMD. Do you have any suggestions?
    I tried to use some artifacts such as drawing images of the water box,
    but it didn't work.
    Este mensaje fue enviado por:

[ Dr Nicolas Sapay ]
Institut de Biologie et Chimie des Protéines
CNRS - Claude Bernard University, IFR128 Bioscience Lyon-Gerland
Phone: +33 (0)4 72 72 26 46 - Fax: +33 (0)4 72 72 26 46
Address: 7 passage du Vercors, 69367-F Lyon cedex 07, France