[chimerax-users] matchmaker question: restrict moving parts to selection?

[Elaine Meng]

Hi Matthias,
The "align" command has a similar option ("move atoms") but "matchmaker" does not, currently.  To move that set of residues separately with matchmaker, you'd have to make it a separate model beforehand.  Ways to do that:

(A) manually (by text-editing) create separate PDB files for the parts you want to move separately before opening them in ChimeraX
(B) open the alphafold model multiple times and delete different parts from the different copies
(C) use the "split" command with its "atoms" option, e.g.

split #1 atoms #1:100-200

<https://rbvi.ucsf.edu/chimerax/docs/user/commands/split.html>

Splitting would create two submodels, e.g. #1.1 with the specified atoms and #1.2 with the remainder.  Check the Model Panel to see what the numbers are in your case, since it might be different. Then specify the desired submodel (or model, if you used methods A or B instead) in the matchmaker command, e.g.

mm #1.1 to #2:100-200

If you need all of the original #1 atoms in one model, you could then combine them back together:

combine #1.1-2 close true model #1

<https://rbvi.ucsf.edu/chimerax/docs/user/commands/combine.html> 

Unfortunately that will use a new chain ID for those residues and there won't be peptide bonds between the different segments, so you'd still have to resort to manual editing to seamlessly incorporate those residues into their original chain.

We realize this is quite a rigamarole compared to a single option in the matchmaker command, so I'll make a ticket requesting it as a new feature.  I hope this helps,
Elaine
-----
Elaine C. Meng, Ph.D.                       
UCSF Chimera(X) team
Department of Pharmaceutical Chemistry
University of California, San Francisco

> On Apr 15, 2022, at 10:22 AM, Vorländer,Matthias Kopano via ChimeraX-users <chimerax-users at cgl.ucsf.edu> wrote:
> 
> Hi everyone, 
> 
> Is it possible to align parts of a model to another while only moving the specified residues but not the rest of the molecule? A use case would be having an alphafold model from one species where the individual domains are correctly predicted, but don't agree with the relative orientation of the domains observed in a reference structure from another species. I envision this similar to the  "fitmap" option "moveWholeMolecule false". 
> 
> Example:
> 
> Mm #1:100-200 to #2/100-200 MoveWholeMolecule false 
> 
> This should leave the residues 1-99 in model #1 unchanged. 
> 
> I hope my question is clear. Thanks a lot in advance, 
> 
> Matthias