[chimerax-users] Multichain comparative modelling

[Elaine Meng]

Hi Kay,
I somewhat doubt that comparative modeling is the way to solve this problem.  Consider:

(1) modeling a multimer requires a multimeric template with same stoichiometry:
<https://rbvi.ucsf.edu/chimerax/docs/user/tools/modeller.html>
Only one of your structures has both proteins, so only that structure could be used as a template.  Alternatively you could try modeling only the monomer of the 2-domain protein using both structures as templates, but the process may change the relative orientations of the domains that you say were already in the correct relative orientation, as it will not take the second protein structure into account.

(2) modeling a heteromultimer requires a separate sequence window for each different chain:
 <https://rbvi.ucsf.edu/chimerax/docs/user/tools/modeller.html>
 It sounds like the structures you have are already for the same proteins you are interested in (identical sequences, not homologs from a different species or anything like that) so you wouldn't even need an alignment, just the single sequence itself.  For heteromultimer modeling, could simply open the structure that contains both proteins, then show both of its sequences (menu: Tools... Sequence... Show Sequence Viewer), which would be in two different sequence windows.... but as mentioned above, heteromultimer modeling may not make sense in your situation.  For monomer modeling, you could just open the single sequence of that one protein with two domains (same menu entry, just choose that one chain as the sequence to view), but I'm not sure that makes sense in your situation either.

Instead you might just try manually editing the PDB files (i.e. using a text editor) to substitute the helical loop conformation into the structure that already has both domains.  If the two structures aren't already superimposed appropriately to do this, you would first superimpose them (e.g. using Matchmaker, menu: Tools... Structure Analysis... Matchmaker), then save PDB of one of the structures  choosing the option to save "relative to" the other structure, then use this newly saved PDB for the text-editing part.  You may need to do some renumbering after that, e.g. open your newly edited PDB, then use menu: Tools... Structure Editing... Renumber Residues
<https://rbvi.ucsf.edu/chimerax/docs/user/tools/renumber.html>

Another possibility is to try to use AlphaFold to predict the multimer (menu: Tools... Structure Prediction... AlphaFold, see the prediction option) but this may not (a) be feasible depending on the sizes of these two proteins, or (b) give the expected/desired result.
<https://rbvi.ucsf.edu/chimerax/docs/user/tools/alphafold.html>

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 4, 2023, at 10:44 AM, Prof. Dr. Kay-Eberhard Gottschalk via ChimeraX-users <chimerax-users at cgl.ucsf.edu> wrote:
> 
> Hi There,
>  
> I am interested  in a complex formed by a protein with two domains connected by a long loop in complex with a second Protein, that binds this loop, causing it to form a short helix. I have two pdb stuctures: One with the two domains in the correct relative domain orientation, connected by an unstructured loop, and one with the part of the loop forming a helix in complex with a second Protein. I am sure that comparative modelling can do the trick, but somehow I dont understand the workflow. Do I need do load both templates into chimeraX, oben sequence file conatining both Proteins and load that also, if so in which format? I went over the Manual, but somehow it is not clear to me.
>  
> Thanks a lot for your help!
> Best,
> Kay
>