Re: Using PBC tools for a cubic and a rectangular box

From: Giacomo Fiorin (giacomo.fiorin_at_gmail.com)
Date: Tue Mar 27 2018 - 08:35:54 CDT

On Tue, Mar 27, 2018 at 9:09 AM, Brian Radak <brian.radak_at_gmail.com> wrote:

> Hi Monika,
>
> The usage of pbctools is a little more complicated than what you are
> doing. I recommend looking at the plugin documentation (
> https://www-s.ks.uiuc.edu/Research/vmd/plugins/pbctools/)
>
> Short explanation - pbctools does not know anything about your box
> dimensions (which MUST be orthorhombic) until you tell it. That information
> is NOT extracted from the DCD frames, but can be read from an XST
> trajectory.
>

I'm not sure that's the problem... "pbc get -all" may tell whether the
correct unit cell is being read for each DCD frame.

Monika, I would definitely try to visualize the trajectory in VMD after
wrapping.

At times I found pbc unwrap to give incorrect results (i.e. a few waters
had bonds longer than one unit cell). In those cases, qunwrap gave correct
results, and was also much faster:
https://github.com/jhenin/qwrap
If the system topology is set in a way that VMD sees each molecule as a
distinct "fragment", I would recommend:
qunwrap compound fragment

This worked well for me for polymer simulations in LAMMPS, which unlike
NAMD wraps every single atoms in the original unit cell (i.e. breaking
bonds).

Giacomo

>
> This will assign your fixed box size to all frames, *then* you can unwrap--000000000000a5ce1b056864fa83--

This archive was generated by hypermail 2.1.6 : Wed Sep 18 2019 - 23:19:47 CDT