Re: [AMBER] AMBER Digest, Vol 989, Issue 1

From: Jason Swails <>
Date: Mon, 29 Sep 2014 08:06:25 -0400

On Mon, 2014-09-29 at 00:42 -0700, amirhossein taghavi wrote:
> Hi,
> I removed the noimage from script and added the parmbox info plus the
> angles as you said and the problem is solved.Thanks a lot!
> May I ask what do you mean by 'corner case' and how can I learn more
> about the algorithms that cpptraj uses to analyze bonds,dihedrals
> and ....

In most cases, either the topology file, the trajectory file, or both
will give cpptraj enough information to determine what the shape of the
unit cell is. Without the shape (i.e., the angles between the unit cell
basis vectors), you have no way of determining the periodic images of
any particle.

It is unusual for files fed to cpptraj to contain _only_ the lengths of
the periodic cells and not the angles. Since this isn't enough
information to perform imaging, cpptraj pretends there is no box (this
is the "corner case"). The algorithm for computing distances and angles
are quite simple -- the same that you learn in basic geometry. When
unit cell vectors are defined, cpptraj will compute distances using the
minimum image convention (i.e., find the image of particle 2 that is
closest to the chosen image of particle 1 and compute that distance).


Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Mon Sep 29 2014 - 05:30:03 PDT
Custom Search