[Chimera-users] Fitting ligands/substrates to protein structure

Elaine Meng meng at cgl.ucsf.edu
Thu Dec 5 15:27:35 PST 2019

Hi Jason,
Short answer is that it is probably a more challenging task than you appreciate.  This is a whole field of computational chemistry: molecular docking given structures of the “receptor” and the “ligand.”  Consider that both molecules are flexible and you may not know what conformation is in the complex, even if you have biochemical knowledge of which amino acid residues of the protein are involved.  There are probably >50 programs that do docking.  

Chimera itself does not do docking.  It does have an interface to a web service run by a different lab that uses the Autodock Vina program for molecular docking.  However, a shared resource, this web service only allows very little sampling of space and conformations.  Thus the results should be viewed with a large dose of skepticism, and you would have to consider carefully whether they are consistent with any other biochemical knowledge (e.g. catalytic residues, mutations known to disrupt binding, etc.).

Here is the help page for this interface in Chimera, which can also be seen by starting that tool (menu: Tools… Surface/Binding Analysis… Autodock Vina) and clicking its Help button:

Another approach is to manually position the ligand yourself, by deactivating (freezing in place) the receptor or ligand and moving the other with the mouse, rotating bonds, etc., to make it consistent with any biochemical knowledge, but this can be an extremely difficult task.

Activating/deactivating models for motion and using the mouse:

Rotating bonds:

The images you’ve seen might be from experimentally determined complexes (i.e. the crystal structure of the protein already contains the ligand), or from running one of the many separate docking programs and then displaying the results in Chimera using its ViewDock tool:

As an example, see tutorials for the DOCK6.9 docking program:

I hope this clarifies the situation,
Elaine C. Meng, Ph.D.
UCSF Chimera(X) team
Department of Pharmaceutical Chemistry
University of California, San Francisco

> On Dec 5, 2019, at 12:28 PM, Jason Jeffrey Rodencal <rodencal at stanford.edu> wrote:
> Hello chimera users!
> I've been browsing the various tutorials for chimera and I've seen a lot of amazing models of proteins bound to DNA, ligands, etc. I have a pdb file for my protein of interest and its substrate, and I have the active site of the protein. I was wondering if there's any way of predicting how the substrate fits into the catalytic cleft. Is there some easy code or application that can do this in chimera or is this a far more challenging task than I appreciate?
> Thank you!
> Jason Rodencal
> Dixon Lab
> Stanford University

More information about the Chimera-users mailing list