From: John Stone (johns_at_ks.uiuc.edu)
Date: Fri Jun 11 2004 - 09:04:43 CDT

Hi Jerome,
  Which version of VMD are you using?
We've had a couple of implementations for the RMS alignment code, so knowing
which version you're using would tell me which alignment code was in use.
The older versions performed less well than the newest version.
However, that said, I also believe there there are numerical circumstances
that can cause many of the implementations derived from the Kabsch algorithm
to fail to converge within a reasonable number of iterations, which may explain
why yours worked on the second attempt. I have seen some structures that
don't align well even with repeated attempts.

  John

On Fri, Jun 11, 2004 at 09:48:49AM +0200, Jérôme Hénin wrote:
> Hi all,
>
> I have met an unexpected behaviour using the "measure fit" command, and I'd
> like to know if anyone has had a similar experience.
>
> When computing the rmsd along a trajectory, some frames were not properly
> aligned with the reference system, which resulted in jumps in the measured
> rmsd. By repeating the fitting procedure, the alignement problem was
> eventually corrected.
>
> Is this to be considered a bug of the alignement algorithm?
>
> Jerome

-- 
NIH Resource for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
Email: johns_at_ks.uiuc.edu                 Phone: 217-244-3349              
  WWW: http://www.ks.uiuc.edu/~johns/      Fax: 217-244-6078