From: Axel Kohlmeyer (
Date: Tue Jan 09 2007 - 14:13:08 CST

On Tue, 9 Jan 2007, Nitin Bhardwaj wrote:

NB> Dear VMD/NAMD users,

dear nitin,

first off, please try to avoid cross-posting to multiple lists.
it is very likely that people are not subscribed to both lists
so one has to manually edit the addresses on a reply to avoid
that the mail gets bounced. thanks.

NB> I saw that during my simulation of a sigle lipid in a water box the
NB> lipid moves out of the box. The water box seems to divide into 8 boxes. From

how many water molecules do you have in that box. i'd guess very few,
so what you are seeing is basically the 'hole' left by the lipid being
replaced by a periodic image.

NB> the mail archive I learnt that this is a problem of the VMD graphics and if

i beg to disagree. this is not a problem of the VMD graphics, this is
just how your trajectory is proceeding. depending on whether you
instruct your MD program to wrap coordinates back into the original
box or not you'll either see molecule jumping in and out of the original
simulation cell and being replaced by one of its periodic images or
the whole cell 'spreading' out. for the calculation of the non-bonded
interactions this PBC wrapping has to be regardless and is usually
done on a per-chargegroup basis in bio-simulations. however for
convenient analysis the wrapping in the output it done on a per-molecule
basis and whether a molecule 'jumps out of the box' depends on the
reference point that decides which periodic image to use (frequently
the first atom in a residue). for more details, please have a look
at some of the MD text books that also cover implementation details,
e.g. allen & tildesley or frenkel & smit.

NB> I added more periodic boxes, the lipid molecule would appear in the water
NB> box.
NB> I was just wondering why does this happen in the first place and what does
NB> it mean. Does it mean that something is wrong with my system/parameters of
NB> conf file? Also if I select "water within 10 of lipid", the water selected
NB> is somewhere else in space and is no where near the lipid. Is it normal?

VMD in its current incarnation has no real knowledge of periodic images,
so any selection or analysis totally disregards them. the periodic
display of representations in a way, that it works for displaying
purposes only. i'm certain that john can confirm that a change of
that has been on the TODO list for a long time, but it sounds easier
than it is to implement, especially if one wants to do it in a way
that the normal case (no PBC) remains as fast as it is now and that
there is no additional memory needed unless you turn on periodic

if you want to do a proper analysis, you'd first have to have
to do some kind realigning the whole trajecory using some
kind of RMS or COM fit on the lipid molecule and then PBC
wrap the waters.


NB> Thanks a lot in advance,
NB> rgds,
NB> Nitin

Axel Kohlmeyer
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
If you make something idiot-proof, the universe creates a better idiot.