The Molecular Graphics Laboratory Forum

Dear all,

I am a new user of AutoDock; I have a few questions: one about how partial charges are calculated, for ligand and macromolecule, during the preparation of the .pdbqt files; one regarding the issue of non integral charges on protein residues or ligands when partial charges are calculated and the last one regarding atoms in .pdbqt-protein files with 0.000 partial charge.

Question A) As I read in the tutorials the usual procedure to prepare .pdbqt files for macromolecule and ligand is to use AutoDockTools and do the following:


for macromolecule:

1)Edit | Hydrogens | add all hydrogens

2)Grid | macromolecule | choose


...and AutoDockTools prepares the macromolecule automatically: partial charge calculation, merging of non-polar hydrogens, atom type definition...
During this automatic procedure are the partial charges calculated using gasteiger method or kollman? Isn't kollman the prefered method for proteins?

for ligands :

1)Add all hydrogens to the ligand.pdb file

2) Ligand |Input |open ligand.pdb

......and AutoDockTools prepares the ligand automatically: partial charge calculation, merging of non-polar hydrogens, atom type definition, torsion tree...During this automatic procedure the partial charges are calculated using the gasteiger method; so gasteiger is the prefered method for partial charge calculation for ligands which are not peptides?

Question B1) Recently I prepared the .pdbqt file for a protein doing the usual:

1)Edit | Hydrogens | add all hydrogens

2)Grid | macromolecule | choose

Unfortunately I received a Warning :

###########################################
"Non integral charges on Protein"
"correct residues"

MET1
LYS327
SER340
and more...

"Charges should be corrected in written output file"
###########################################

What can cause this kind of problem? Anyway to solve it I manually did:

1) Edit | Hydrogens | add all hydrogens
2) Edit | Charges | Add Kollman charges
3) Edit | Charges | Check totals on residues | spread Charge Deficit over all atoms in residue
4) Edit | merge non polar hydrogens
5) Atoms | assign AD4 type
6) File | save |write pdbqt

Do you think that this procedure is correct? should I add Gasteiger instead of Kollman partial charges? How important can these non integral charges be for the docking?

QuestionB2) Recently I prepared the .pdbqt file of a ligand containing a phosphate group, a carboxyl group and a protonated pyridine ( actually it is the pyridoxal phosphate cofactor bound to carbidopa (inhibitor) which I want to dock in the active site of the PLP-dipendent enzyme) with the procedure:

1)Add all hydrogens to the ligand.pdb file

2) Ligand |Input |open ligand.pdb

unfortunately I received the message:

###########################################
setup_PLPcarbidopa:
added gasteiger charges
merged 15 non-polar hydrogens
found 11 aromatic ...
...
total charge error = 0.4999
###########################################

What can cause this "total charge error"? Anyway to solve it I manually did:

1) Edit | Hydrogens | add all hydrogens
2) Edit | Charges | Add gasteiger charges
3) Edit | Charges | Check totals on residues | spread Charge Deficit over all atoms in residue
4) Edit | merge non polar hydrogens
5) Atoms | assign AD4 type
6) File | save |write pdbqt

Do you think that this procedure is correct? How important can these non integral charges be for the docking?

Question C)

I noticed that some atoms in the macromolecule pdbqt file have partial charge 0.000; I read in some tutorial that this can happen when AutoDock does not recognize the name of the atom/parent residue (and therefore can't retrieve the partial charge for that atom...); in my case this happens only for CB atom of GLU residues and CE2 atom of TRP residues...? Is there a way to find the partial charge normally given to these atoms?

Thank you very much for any help that you can give

kind regards

Frederick Daidone

Dear colleagues,

I have not the same, but similar problem with adding charges to ligands, organic molecules (porphyrin derivatives), which deliberately should have positive charges from +2 to +8 depending on pH value. Unfortunately, no .pdbqt file created using Autoodck (mgl-tools) shows any charge when using Edit\Charges\compute Gasteiger or Edit\Charges\Add Kollman Charges. It always shows 0.0000 or something like 0.0002. I tried to change Nitrogen atom 'types' and it did not help. After docking procedure using such 'uncharged' molecules by vina I am terrified seeng positively charged ligand's residue in close proximity to another positive group of the macromolecule.
Can somebody help ? Thank you in advance.
Ulysse

Dear colleagues,

I have figured out how to assign proper charge to N containing residues, and finally obtained a ligand with proper total gasteiger charge of +4. Unfortunately, all my efforts were in vain. When I save this ligand with +4 total charge on the residues as a .pdbqt file and then reopen it again it says: calculated gasteger charge = zero ! I use AutoDockTools 1.5.4 revision 30. Please help. I am disastered, giving up ((((((((((((((((((((((((((((((((((((((((((

Ulysse

Hi Ulysse,
Your message suggests that after adjusting the charges in your ligand file, you are 'designating your ligand' in ADT by "reopening" it using Ligand->Input->Open which re-formats it thus clobbering your charges.

Instead, since you already have the ligand file you want to use, at this point go ahead and prepare the rest of the input for your docking experiment.
-To prepare the gpf file use 'Set Map Types->Choose Ligand...' or 'Set Map Types->Open Ligand' in the Grid menu.
-To prepare the dpf file use 'Ligand->Choose...' or 'Ligand->Open...' in the Docking menu.

Dear Ruth,

Thank you very much for the respond. The problem is that I do not use ADT for docking, I use vina instead, which is more efficient and time saving docking program. And vina works from command line where ligand and macromolecule .pdbqt files should be specified. So I have to create . pdbqt file beforehand, save it and then reopen by vina. As far as I know the Autodock Vina is he best product (developed by Scripps Institute) for docking and screening.
Best regards,
Ulysse

Hi Ulysse,
ADT is the same as AutoDockTools: the python-based graphical-user-interface for preparing input for vina and for autodock4.

Your original message points out that you "reopen"ed your ligand in AutoDockTools 1.5.4 revision 30:
"When I save this ligand with +4 total charge on the residues as a .pdbqt file and then reopen it again it says: calculated gasteger charge = zero ! I use AutoDockTools 1.5.4 revision 30"

How and why did you reopen it in AutoDockTools?
Ruth

Doesn't Vina completely ignore charges anyway?

Dear Ruth,

I wrote you a big message a week ago, but it was not posted. Probably, moderator decided it's too long.
Shortly, I follow instructions of videotutorial by Oleg Trott, one of the vina authors.
Regarding the question " Why I reopen pdbqt file" . I reopen it to be sure the partitial charges were assigned rightly, and I sow they were not.
Seems, Douglas is partially right. In the Autodock Vina FAQs I found the following.

http://vina.scripps.edu/manual.html#faq
Why don't the results change when I change the partial charges?

AutoDock Vina ignores the user-supplied partial charges. It has its own way of dealing with the electrostatic interactions through the hydrophobic and the hydrogen bonding terms. See the original publication [*] for details of the scoring function.

Then, I do not understand why vina requires pdbqt files as input, but not simply pdb ?

Thank you, all who took part in the discussion,
Ulysse

Because PDB files don't specify rotatable torsions or atom types.