VMD-L Mailing List

From: Luis Gracia (lug2002_at_med.cornell.edu)
Date: Fri Nov 19 2004 - 15:03:37 CST

Hi Gavin,

after looking at your structures, I think that the problem arises
because the 3 monomers in the bad aligned "corners" are in different
order than in the other corners.
If I align your structures in reference to cnr1 using 'all' or 'resid
395 to 637' I get the same results as you, with cnr3, cnr5 and cnr7
misaligned. But using 'resid 395 to 637 and (chain G I)' only for
molecules cnr1 and cnr3 (to make it easier) aligns them perfectly. I
haven't try with the other two corners, but I guess is the same problem.

By the way I am not using the measure fit command, but the rmsdtt plugin
(the rmsd tool plugin should give the same results).

Hope this helps,

Luis 

--
Luis Gracia, PhD
Department of Physiology & Biophysics
Weill Medical College of Cornell University
1300 York Avenue, Box 75
New York, NY 10021
Tel: (212) 746-6375
Fax: (212) 746-8690
lug2002_at_med.cornell.edu
Gavin Murphy said the following on 11/18/04 23:37:
> Hello,
>   I am working on the cubic core of pyruvate dehydrogenase.  It has 8 
> corners, each made of 3 monomers.  I am trying to align 7 of the 8 
> corners to one particular corner.  I get bad alignments with three of 
> the corners, and I believe that it has to do with symmetry.  They are 
> the corners that are diagnolly opposite one another on any given face. 
> All the other corners align well.  I am using the
> measure fit command.  Is there a bug in the software, or am I missing 
> something?  I could give out the pdb files to anyone who wants to 
> reproduce it.
> 
> Thank you,
> Gavin Murphy