[Chimera-users] Match -> Align used with a pair of homodimers

[Elaine Meng]

Hi Eric,
The steps in my previous message included Matchmaker fitting using  
only one monomer of each structure.  In principle, you might want to  
do the initial fitting using both monomers (maybe that's what you were  
saying in your question and I didn't understand).  However, in this  
case it doesn't make much difference because fitting a monomer  
superimposes the whole dimer pretty well.

To use both chains in Matchmaker, you would choose the "Chain pairing"  
option "Specific chain(s) in reference structure with specific  
chain(s) in match structure."  Then you can specifically the pairing  
of A with A and B with B (or A/B and B/A, equally valid):  drag to  
choose both chains of 1nox dimer on the left, and then on the right,  
use the pulldowns to specify which 1ds7 chain should be paired with  
each.

Otherwise using Matchmaker and Match->Align defaults:
A/A,B/B pairing - Matchmaker 188 atom pairs 1.139 ang. RMSD,  
alignments from Match->Align ~23 and 21% ID.
A/B,B/A pairing - Matchmaker 188 atom pairs 1.139 ang. RMSD,  
alignments from Match->Align ~23 and 21% ID.

Compare to initial fit on only one pair (reported in previous mail) of  
101 atom pairs 1.026 ang. RMSD, but essentially the same results from  
Match->Align.
Best,
Elaine
-----
Elaine C. Meng, Ph.D.                          meng at cgl.ucsf.edu
UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab
Department of Pharmaceutical Chemistry
University of California, San Francisco
                      http://www.cgl.ucsf.edu/home/meng/index.html



On Feb 23, 2009, at 5:51 PM, Elaine Meng wrote:

> Hi Eric,
> A couple of us independently tried the process and did not encounter
> the problem.  Using default matchmaker and match->align parameters,
> the two alignments from match->align had 23% and 21% ID.  However, I
> had a theory of where you might have gone astray.  Here are the steps
> of the process:
>
> (1) generate 1nox dimer as a single model.  Note this is much easier
> in newer versions of Chimera (1.3+):  after creating the dimer using
> BIOMT, you can then use the "combine" command or the "copy/combine"
> action in the Model Panel to merge the two monomer models into one
> dimer model with chains A and B.  No hand-editing!
>
> (2) have 1nox dimer and 1ds7 open in Chimera, use Matchmaker with
> default settings.  This gives a sequence alignment of 1ds7 chain B
> and 1nox-dimer chain A with ~26% ID, as you said, and a pretty good-
> looking superposition: final iteration 101 atom pairs, 1.026
> angstroms RMSD.
>
> (3) in Match->Align, you would choose 1ds7 chain B and 1nox-dimer
> chain A as one pair, Apply to get sequence alignment with 23% ID
> (default parameters in Match->Align).  Choosing 1ds7 chain A and 1nox-
> dimer chain B as another pair, clicking Apply gives another sequence
> alignment with 21% ID.
>
> My theory is that maybe you chose A and A as one pair, B and B as
> another pair whereas the superposition really has each A chain on top
> of a B chain.
>
> I hope this helps,
> Elaine
>
>
> On Feb 23, 2009, at 4:41 PM, E. Merkley wrote:
>
>> Hello chimera friends:
>>
>> I am trying to get a structure-based sequence alignment for a pair of
>> homologous proteins that are homodimers.  If I proceed as described
>> in the
>> tutorial and in the 2006 BMC Bioinformatics paper, by first using
>> Matchmaker, then using Match -> align, the Match -> Align fails to
>> produce
>> a reasonable alignment.  The sequence identity from the Matchmaker
>> output
>> is 26.5%, but only 1.0 % from the Match -> Align step. If I delete  
>> the
>> second monomer from each protein, Match -> Align works quite well.
>> However, it seems like this alignment won't take into account any
>> differences in the tertiary structure between the two proteins.
>> That is,
>> I think I want the global alignment of the whole dimer for the
>> structural
>> alignment, since the active sites is are at the interface.  Any
>> suggestions?  I'm using version 1.2540, and my two proteins are PDB
>> codes
>> 1ds7 and 1nox (1nox dimer built from BIOMT matrix in Chimera and
>> hand-edited to be 1 model with A and B chains).
>>
>> Thanks yet again,
>> Eric
>>
> _______________________________________________
> Chimera-users mailing list
> Chimera-users at cgl.ucsf.edu
> http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users