how to add Cadmium ion to structure and do the refinement

From: Lu Yu
Date: 28 November 2011 20:16


Hi ALL,

This is the first structure I am working with.
As the title said, I was stuck in the middle of model building using coot, and I found a big positive peak for Cadmium ion. The coot program doesn't *know* the cadmium ion when I add a Cd atom there (under the "place atom at pointer" menu--> "Other" ), and the refinement program doesn't *know* how to refine this (failed when I refine the whole structure, and it may need appropriate values as restraints, but I don't know how to add these values ).

So I am wondering any one can give me any suggestion about, how to add cadmium ion in the position and do the refinement.

Thanks in advance,

Lu

----------
From: Roger Rowlett

Lu,

To add a metal ion to a structure, you can use "place atom at pointer" or "Get Monomer...". I usually use the latter. For instructions see http://capsicum.colgate.edu/chwiki/tiki-index.php?page=Model+Building+and+Validation#Adding_ligands_to_a_protein. For cadmium, I think you can enter "CD" for the monomer name (all caps). The documentation for Refmac has a complete list of recognized monomers and their 2-3 letter names.

To refine the metal in Refmac with restrained ligand distances, see http://capsicum.colgate.edu/chwiki/tiki-index.php?page=Model+Refinement#Restraining_metal_ligand_distances_during_refinement_in_REFMAC. Refmac will create a .cif file with the proper restraints when you try to run it the first time (and fails). Include the library (.cif) file when you re-run the job and all will be well.

BTW, if this is a Cd(II) ion, you should really edit columns 76-80 (I think that's right--you'll see the elements listed in the last 4 columns of each line of the pdb file) of the PDB file to read "CD+2" and not "CD  ", although in practice this may not make that much difference. But is is more accurate in terms of X-ray scattering if you indicate the ionic charge.

----------
From: Jim Pflugrath


If you edit the PDB file "by hand", then you should have no problems.  Some folks might cringe at this, but this does work:

1. Like Roger Rowlett said, "place atom at pointer".  You can place a chloride there.  Be sure to add to your molecule and not a new molecule.
2. Save the coordinates to a PDB file.
3. Edit the newly saved PDB file with the Chloride designation with an editor (emacs is great) and change the Chloride to a Cadmium.  What does a Cadmium look like in a PDB file? I don't have a clue, but I would search for "cadmium" at rcsb.org and save a PDB file with a cadmium in it.  Then you can see what a cadmium looks like in the example PDB file and mimic that in your file.

Jim
----------
From: Zhijie Li

Hi Lu,

 

Cadmium(II) seems to be in the standard monomer library in both Coot and CCP4. Thus you do not need to make it by yourself. I guess your problem was due to that you have put a "Cd" instead of "CD" when adding it in Coot. Refmac is case sensitive when identifying the compounds or atoms. All ions need to be in upper case to match their .cif files.

 

Please check the "share\coot\lib\data\monomers\c" and "ccp4-6.2.0\lib\data\monomers\c" folder. You should have a "CD.cif" there. This file is the monomer definition for Cd(II) ions. Please take a look in the file, especially this part:

 

loop_
_chem_comp_atom.comp_id
_chem_comp_atom.atom_id
_chem_comp_atom.type_symbol
_chem_comp_atom.type_energy
_chem_comp_atom.partial_charge
_chem_comp_atom.x
_chem_comp_atom.y
_chem_comp_atom.z
 CD            CD     CD   CD        2.000      0.000    0.000    0.000

 

The 2.000 in the last line corresponds to the "_chem_comp_atom.partial_charge" entry. The 2.000 indicates that this is for a Cd(II) ion. If your Cadmium is Cd(II) then you are OK. No more worries.

 

If you think your Cadmium is Cd(III), then you need to change the partial_charge value to 3.000, and change all the "CD"s in this file to "CD3", then save it as another file, "CD3.cif" (make sure "CD3" is a unique name in the whole library), and use it for your refinement. Also in this case you need to put "CD3"s in your PDB file instead of "CD".

 

This is a simple case so you can edit by hand. Please take a look in the "FE.cif" and "FE2.cif" in the "f" folder to see how such cases should be handled. Also please note that the FE1.cif and FE3.cif are not for iron at all. This is an example showing that we should always have a look in the cif files to make sure the descriptions are correct for our ligand. If not we need to find the right one or make our own.

 

Zhijie

----------
From: Zhijie Li

Corrections of my previous email:

 

According to wikipedia:

"Although cadmium usually has an oxidation state of +2, it also exists in the +1 state."

So there is probably no Cd(III) for us to worry about.

----------
From: Eleanor Dodson


Hasanyone answered this?

Possibilities
1) COOt by default may have written your CD to another coordinate file - usually called PointerAtoms.pdb

2) You got the atom into the pdb you want to refine but it is wrongly labelled.

You need
ATOM    471  N   ASP A  81      23.896   6.780   5.956  1.00 60.81..

the atom label moved to the left, and the residue to be called CD and the atom type (last entry on ATOM line) to be CD
ATOM    471 CD   CD  A  81      23.896   6.780   5.956  1.00 60.81..

The residue number must be different from any existing residue..

The REFMAC log will trell you what it disliked, but not very well marked..

----------
From: Lu Yu

Hi,
Thanks for your advice!
My problem has been solved, coot has cadmium ion (which I didn't know at first), and the reason that it didn't work at first is I put Cd instead of CD (should be *upper case*) under the "place atom at pointer" menu, so the refmac and coot don't recognize it. Now I learned new things!

Many thanks!

Lu