[Chimera-users] reflection of docking poses with chimera

[Elaine Meng]

On Apr 18, 2009, at 9:43 AM, Antonios Kolocouris wrote:
> Hi Elaine,
> I have obtained the results of the docking claculations of some  
> compounds
> with a symmetric - by 180 degrees - receptor. To obtain a  
> consistent set of
> structures for further 3D QSAR I need to know how I can reflect a  
> docking pose by 180 degrees
> vertically and horizontally with Chimera.
> My kind regards,
> Antonios
> ---------------------------------------------------------------------- 
> ---------------------
> Antonios Kolocouris, Assistant Professor of Medicinal Chemsitry
> Department of Pharmacy, Univ. of Athens, Greece
>

Hi Antonios,
The main issue is figuring out the proper transformation to  
superimpose one binding site on the other.  It will not exactly be a  
reflection (you will still have the same enantiomers) but a rotation.

You would need a recent daily build for what I describe.

If you are certain only rotations about the exact horizontal and  
vertical axis are needed, and you know what the center of rotation  
should be,  use the "turn" command with specified axis and center of  
rotation to move the ligand model(s).  The horizontal axis is x, the  
vertical axis is y.
<http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/turn.html>

The rest of this message addresses the more general case where the  
axis of symmetry and thus the proper transformation is not known:

-- To find out the transformation,

(A) put the two parts containing the two binding sites in different  
models.  This could be done by manually text-editing your PDB file to  
create two files that can then be read into Chimera.  Or, if they are  
already two different chains, you can use the "split" command to  
divide the already open model into separate models.
<http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/split.html>

(B) superimpose the two binding sites.  It could be done with  
matchmaker, or match, or manually.  Which is best depends on the  
situation.  These methods are discussed and there are links to their  
manual pages here:
<http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/superposition.html>

(C) find out the relative transformation of the two models using the  
"measure" command, for example to report the transformation of model  
#0.2 relative to #0.1:

measure rotation #0.1 #0.2

the transformation is sent to the Reply Log (under Favorites menu),  
both as a matrix and as an axis of rotation and point on the axis, as  
described in the manual page:
<http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/measure.html>

-- To apply the transformation to the ligands, I would first use the  
command "reset" (without arguments) to put everything back into their  
untransformed positions and then either

(A) use the "turn" command with specified axis and center of rotation  
to move the ligand model(s)
- or -
(B) save the matrix from the Reply Log to a file and then apply it  
with the "matrixset" command

I'm not sure if the "reset" is necessary (some of these  
transformation issues confuse me), but that's what I would try first.

I recommend sending questions to chimera-users at cgl.ucsf.edu rather  
than to me directly, to be certain they will not fall through the  
cracks.

I hope this helps,
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