On 08/03/2018 17:55, Thomas, Leonard M. wrote:
> Hello All,
Hello Thomas,
>
> We have run into a problem recently when working with cif files for ligands that have been either soaked or
> cocrystallized with the proteins we are looking at. Most of the ligands cif files have been used previously
> and while not in the CCP4 library database are available from the pdb ligand database. Now when we go to
> use them, specifically in coot after refinement in Refmac we get an error saying there are no restraints for
> the ligand.
Coot may be telling the truth. It depends on what you mean by "used previously" - if you mean that you can
use these cif files in Refmac without problem then I guess that there's a problem with Coot. As you know,
cif files can contain data, models, chemical definition or (more pertinently) ligand restraints. Coot
expects cif files with ligand restraints to have a particular format. After reading a cif file that you type
as a dictionary as you read it into Coot, you can see what Coot was able to parse from it by looking in Edit
-> Restraints. Restraints contain blocks including _chem_comp_angle, _chem_comp_tor, _chem_comp_plan.
Chemistry definition cifs don't have these.
> We have tried reading the cif file in before and after reading the pdb file in.
Either will do. But Coot has special rules for ligand named LIG, INH, and DRG - ligands of these types don't
force an auto-load of the Refmac Monomer Library.
>
> We can put the molecule/ligand and cif file into Phenix and all seems well with coot but when moving back to
> CCP4/coot we still have the problem.
Phenix uses Coot's API to send it restraints information on your behalf.
> Both CCP4 and Phenix are using the same version of coot so I don’t
> think it is a version problem with coot.
>
I agree.
Paul.
|