Hello Simon, I had that once too with a cif-file from Prodrg. In my case the problem was that in the cif-dictionary two different torsion angles where defined for the same four atoms. Deleting one of the definitions of the torsion angles in the cif file solved the problem. This also solved the problem of the ligand falling appart in real space refinement in coot. Sabine Folmer Fredslund wrote:

Dear Simon

2008/8/7 Simon Kolstoe :

...

phenix.refine keeps on stopping with the error:

AssertionError: Conflicting dihedral restraints: tor id: var_002 1. definition from: residue: DRG, conformer A 2. definition from: residue: DRG, conformer A atoms: " CB DRG A 500 " " OG1 DRG A 500 " " PAL DRG A 500 " " OAF DRG A 500 "

It is for a phospho-threonine residue whose dictionary I made in PRODRG. I've tried all the different charge states of the phosphate however phenix gives the same error every time. Any ideas?

I guess you could try and use the naming from a phospho-threonine residue? try having a look at HIC-Up, a phosphothreonine is tpo

http://xray.bmc.uu.se/hicup/TPO/index.html or http://xray.bmc.uu.se/hicup/ to start a search

HTH

Regards, Folmer

...

Thanks,

Simon _______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb

Thanks everyone. I had already tried the hiccup TPO file which hadn't worked either, but elbow seems to work well. Simon On 7 Aug 2008, at 16:30, Pavel Afonine wrote:

Hi Simon,

did you try to use eLBOW to generate a cif file? If you use PHENIX, then using eLBOW is an obvious choice to create cif files for your ligands:

phenix.elbow model.pdb --residue=XXX

or

phenix.elbow model.pdb --do-all

Pavel.

On 8/7/2008 7:40 AM, Simon Kolstoe wrote:

...

phenix.refine keeps on stopping with the error:

AssertionError: Conflicting dihedral restraints: tor id: var_002 1. definition from: residue: DRG, conformer A 2. definition from: residue: DRG, conformer A atoms: " CB DRG A 500 " " OG1 DRG A 500 " " PAL DRG A 500 " " OAF DRG A 500 "

It is for a phospho-threonine residue whose dictionary I made in PRODRG. I've tried all the different charge states of the phosphate however phenix gives the same error every time. Any ideas?

Thanks,

Simon _______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb

Simon, I am not familiar with the mechanism behind cif-links, but when I refined NAG-aa complexes I just defined the relevant bonds and angles under geometry_restraints.edits.bond or .angle . You need to select individual atoms in this case though. I can dig out an old .def file and send it to you if you get stuck. Carsten

-----Original Message----- From: [email protected] [mailto:[email protected]]On Behalf Of Simon Kolstoe Sent: Friday, August 08, 2008 9:46 AM To: PHENIX user mailing list Subject: [phenixbb] Another dictionary question...

Sorry for one more question:

I am now trying to build in some sugar NAG residues that are N-linked to an Asn. I've used the parameter file below, and the first time phenx.refine ran perfectly. However after a bit of rebuilding I get the following error:

result.delete_atom_in_place(mod_atom.atom_id) File "/usr/local/phenix-1.3-rc4/mmtbx/mmtbx/monomer_library/ cif_types.py", line 244, in delete_atom_in_place raise RuntimeError("delete_atom_in_place: unknown atom_id: %s" % atom_id) RuntimeError: delete_atom_in_place: unknown atom_id: O1

I think this is referring to the O1 atom of the NAG sugar which the program had deleted the first time it ran in order to form the bond with the Asn. Is there any way of stopping phenix trying to delete this atom again?

Thanks,

Simon

My cif_links.param file that is incorporated into my .def file second time round:

refinement.pdb_interpretation.apply_cif_link { data_link = NAG-ASN residue_selection_1 = chain A and resname NAG and resid 207 residue_selection_2 = chain A and resname ASN and resid 32

_______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb

Thanks - Ralph solved my problem. I looked in the cif dictionary and as I had originally added the NAG's using CNS the O1's were already removed. When I generated a dictionary using elbow, as the O1's were not present in the pdb coordinates, they did not get included in the dictionary - and hence the error. Not sure why it ran OK first time round... Simon On 8 Aug 2008, at 17:28, Ralf W. Grosse-Kunstleve wrote:

Hi Simon,

...

RuntimeError: delete_atom_in_place: unknown atom_id: O1

It is trying to delete the atom in the monomer library definition, as opposed to in the PDB file. It has to delete the atom in the monomer each time phenix.refine runs. So that looks OK. I'm guessing you somehow have the link twice. Could you look for a duplicate apply_cif_link block in your inputs? -- But I'm not sure that's it. If not, could you send me your inputs (directly, not to the list)?

Ralf _______________________________________________ phenixbb mailing list [email protected] http://www.phenix-online.org/mailman/listinfo/phenixbb