[Chimera-users] Protonating the protein

[Elaine Meng]

Hi Francesco,

If you know which carboxylates you want to be protonated and are going  
to use AMBER anyway, the most straightforward approach may be to  
simply rename the residues with the appropriate AMBER names (ASH and  
GLH, I believe, for the -COOH forms) and let AMBER take care of it.   
Performing MD will move everything around from the positions chosen by  
some other program anyway.

However, if you are using charges assigned in Chimera for something  
(Dock Prep with AddH, Add Charge,  and write Mol2 for input to DOCK  
scoring grid calculations, for example), you would need to add the  
protons in Chimera before assigning charges.  This could be done by  
selecting the carboxylate oxygens you want to get protonated (no more  
than one per carboxylate group, of course) and changing their atom  
types with the command:
  setattr a idatmType O3 sel
prior to running AddH and Add Charge (alone or as steps in Dock  
Prep).  It is not necessary to rename the residues.  Add Charge will  
then recognize the neutral forms of ASP and GLU and use the  
corresponding partial charges.  I just tested this in the Chimera  
production release, 1.2540 (July 9).

When you switch from one program to another, there are often naming/ 
formatting issues.  Chimera generates hydrogen names to match the  
newer PDB specifications, while others may use older PDB hydrogen  
names or their own naming system (e.g. AMBER).  To get the protonated  
structure from Chimera into AMBER, you might have to rename all the  
hydrogen atoms and those special residues, or maybe just  delete all  
the hydrogens, rename the residues, and have AMBER add hydrogens back  
for its own calculations.

There are some Web servers  that add hydrogens after predicting which  
carboxylates (if any) or other ionizable groups should be protonated.   
Again, there may be formatting and naming issues -  some servers  
output PQR format, which is not quite PDB format. If you are not sure  
which carboxylates should be protonated, you could simply run these  
predictions to help decide which residues to rename for hydrogen  
addition by AMBER.

Here are a few of the many Web servers that can add hydrogens to  
structures:

MolProbity (can add hydrogens with Reduce):
http://molprobity.biochem.duke.edu/

H++
http://biophysics.cs.vt.edu/H++/

PROPKA/PDB2PQR
http://propka.ki.ku.dk/~drogers/

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 Aug 7, 2008, at 3:38 AM, Francesco Pietra wrote:

> Hi:
> I am not saying anything new. I have a (for me) difficult problem of
> protonating a protein homology model (built with Modeller) where there
> are several halogen ion ligands. Actually I succeeded in getting a
> workable model. Chimera loads it and add hydrogens. I could even
> prepare the model for DOCK6, generating spheres.
>
> Now, however (and also because, before docking small ligands with
> DOCK6, I want to dock polypeptides, probably with DOT2) I would like
> to assign specific protonation states to certain carboxylic acids.
> That is, I want to create an overall model where certain facing
> carboxylic acids actually face as "carboxylate-carboxylic acid" or as
> "carboxylic acid-carboxylic acid", or, finally, as
> "carboxylate-carboxylate", the latter also with metal ligand in
> between or not.
>
> Question: to which extent can Chimera aid the task? And which other
> package (pdb2qr, H++, ...) may aid Chimera? Or no other package, but
> simply renaming the said residues in the pdb file and let Amber 10
> (which is the package I'll use for MD, with ff99B ff for the protein)
> do the job of eliminating clashes?
>
> All that shows that I have no experience in selective protonation, and
> the halide ions (in the above said work, I have them simply under
> "ATOM" under the specific chain of the amino acid residues, not
> "HETATM", as the latter caused a lot or problems) create difficulty to
> protonating packages.
>
> Thanks for giving me a general guideline along with best moving.
>
> francesco pietra