From: Jérôme Hénin (
Date: Tue Feb 18 2020 - 11:23:08 CST


You make a good point. qunwrap does implement unwrapping with different
groupings, but it seems the per-atom option should suffice to cover all
cases. At any rate, maybe I shouldn't change something that isn't broken
(...anymore, since I fixed per-atom unwrapping a little while ago :-)


On Mon, 17 Feb 2020 at 01:15, Axel Kohlmeyer <> wrote:

> Giacomo,
> are you sure that you are not mixing up things?
> I thought for pbc unwrap, it was irrelevant what grouping you choose,
> because it would avoid "jumps" for any atom on a per-atom basis.
> So if a molecule is complete and not wrapped in the first frame, it will
> remain so. So get the "complete" molecule, if it is not already, you would
> use pbc join on that first frame, where indeed you have to specify which
> kind of property you want to use to declare an entity "whole".
> The compound also then comes into play, if you want to "wrap back" the
> unwrapped molecules into the unit cell, as with unwrap, they will move and
> spread out.
> In summary, you typically would do a sequence of pbc join (needed only for
> the first frame!, this is a rater slow operation), pbc unwrap and then pbc
> wrap to get a trajectory as was requested by the OP.
> BTW: if the plan is to make an animation, perhaps some additional
> recommendation. Having molecules jump in an out of the box in every step,
> makes for very difficult to watch and "busy" movies. so it can be more
> pleasant to watch, if there is another post-processing step, where you loop
> over the trajectory after the join/unwrap/wrap and do another unwrap, but
> only for small chunks (say 10-20 frames), so that the jumps of molecules
> are less frequent and thus less distracting.
> axel.
> On Sun, Feb 16, 2020 at 2:00 PM Giacomo Fiorin <>
> wrote:
>> Hello Minjung, "pbc unwrap -compound fragment" (or "qunwrap compound
>> fragment" - see link below) would be the way to go, but you should
>> ensure that the fragment IDs are correctly assigned. Verify this with:
>> set sel [atomselect top all]
>> $sel get fragment
>> If you don't get the same fragment ID for all the atoms inside one POPC
>> molecule, you need to either use something else, e.g. "-compound residue"
>> or tell VMD how to get the correct fragment IDs based on how your system is
>> set up and the file format you use.
>> On Sat, Feb 15, 2020 at 7:05 PM Minjung Godfrey <>
>> wrote:
>>> Hi,
>>> I have a bilayer membrane with POPC with which I'm trying to study the
>>> average diffusion rate of individual molecule. Basically, I need to unwrap
>>> all, get the center of mass for each molecule, and calculate the average
>>> diffusion coefficient of all molecules. When I unwrap it, some molecules
>>> get stretched or cut. When it is wrapped, molecules on the boundary get cut
>>> (part of the molecule appears on the other side of the box). "-compound
>>> fragment -all" does not keep the molecules intact either. Is there a way to
>>> keep the individual molecule intact as they jiggle around the edge? How can
>>> I keep it whole and only jump to the other side of the box when, say, half
>>> the molecule drifts across one boundary?
>>> Thank you!
>>> Minjung Godfrey
>> --
>> Giacomo Fiorin
>> Associate Professor of Research, Temple University, Philadelphia, PA
>> Research collaborator, National Institutes of Health, Bethesda, MD
> --
> Dr. Axel Kohlmeyer
> College of Science & Technology, Temple University, Philadelphia PA, USA
> International Centre for Theoretical Physics, Trieste. Italy.