[phenixbb] Glycan Links/Restraints

Damian Ekiert dcekiert at scripps.edu
Tue Mar 1 14:09:10 PST 2011


Engin,

Thanks for your help.

Initially, I tried the order you are suggesting for the links ("inner first, outer second"), but Phenix quits with the following output:

Sorry: Traceback (most recent call last):
  File "/jcsg/software/phenix/phenix-1.7/phenix-1.7-650/cctbx_project/mmtbx/monomer_library/pdb_interpretation.py", line 972, in apply_mod
    mod_mon = self.monomer.apply_mod(mod_mod_id)
  File "/jcsg/software/phenix/phenix-1.7/phenix-1.7-650/cctbx_project/mmtbx/monomer_library/cif_types.py", line 387, in apply_mod
    result.delete_atom_in_place(mod_atom.atom_id)
  File "/jcsg/software/phenix/phenix-1.7/phenix-1.7-650/cctbx_project/mmtbx/monomer_library/cif_types.py", line 304, 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
apply_mod failure:
  pdbres="ASN A  91 "
  comp id: ASN
  mod id: DEL-O1

This sounded to me like it was trying to delete O1 from my Asn instead of the GlcNAc, which I figured meant that I should reverse the order (i.e., outer residue first, inner residue second).  When I did this, phenix completed refinement with no obvious problems in the log file.  I initially took this to mean that all was well, until I looked at my coordinates.

As for glycans in the same chain or different chains, I agree with you completely.  I actually split them into separate chains to try to get around another apparent problem that was cropping up near the top of my log file:

    Chain: "C"
      Number of atoms: 5065
      Number of conformers: 2
      Conformer: "A"
        Number of residues, atoms: 327, 5018
          Unusual residues: {'MAN%DEL-HO3': 1, 'MAN%DEL-HO4%DEL-O1': 1, 'NAG%DEL-O1': 1, 'NAG%DEL-HO4%DEL-O1': 1}
          Classifications: {'undetermined': 4, 'peptide': 323}
          Modifications used: {'DEL-HD22': 1, 'DEL-HO3': 1, 'DEL-O1': 3, 'DEL-HO4': 2}
          Link IDs: {'PTRANS': 18, None: 4, 'TRANS': 304}
            Not linked:
              pdbres="NAG C 401 "
              pdbres="NAG C 402 "
            Not linked:
              pdbres="NAG C 402 "
              pdbres="BMA C 403 "
            Not linked:
              pdbres="BMA C 403 "
              pdbres="MAN C 404 "
          Unresolved non-hydrogen bonds: 1
          Unresolved non-hydrogen angles: 2
          Unresolved non-hydrogen chiralities: 1

These "Not linked" sections went away when I moved the glycans to separate chains, so I figured that Phenix might only allow one polymer type in a given chain (not sure if this is true or not).

As for the monomer cif files, I don't see an major differences between the cifs from Phenix and those from CCP4 that I have used previously for refinement in Refmac, but I will look into this further.

Best,

Damian





On Feb 28, 2011, at 11:44 PM, Engin Özkan wrote:

> Also, I just remembered that Phenix/monomer library may still not have 
> BMA and MAN right. You may have to provide MAN and/or BMA cif files with 
> the correct stereochemistries. Monomer library does not necessarily 
> follow the standards of the PDB chemical component library (but 
> obviously, PDB enforces them).
> 
> Finally, you can always check the .geo file that phenix.refine produces, 
> and you will find all the applied restraints. That should definitively 
> tell you your problem.
> 
> Engin
> 
> On 2/28/11 10:36 PM, Engin Özkan wrote:
>> Dear Damien,
>> 
>> Let me make a guess: You have the BETA1-4 linkages in the wrong order. 
>> If you look in the mon_lib_list.cif file, you can find the line where 
>> it says:
>> BETA1-4  1 O4      2 C1        single       1.439    0.020
>> which tells me that the residue to be defined (residue_selection_1) 
>> ought to be the atom atom connecting with the O4 atom (the inner 
>> glycan) and the second residue is the one with the C1 atom (the outer 
>> glycan). So try switching residue_selection_1 and _2 for your BETA1-4 
>> linkages. I know you are probably thinking BETA1-4 should go 1 to 4, 
>> but that is not how it is defined in the monomer library. By the way, 
>> your NAG-ASN linkages do look correct, and you have the correct 
>> anomers for beta-mannose (BMA) and alpha-mannose (MAN), which is where 
>> most people seem to get stuck.
>> 
>> One stylistic question I have for you is, why are you naming your 
>> protein and glycans with different chain IDs; aren't they actually 
>> covalently linked? There apparently is no right way for numbering 
>> glycan residues, and everybody seems to be do a different thing...
>> 
>> Good luck,
>> Engin
>> 
>> On 2/28/11 6:39 PM, Damian Ekiert wrote:
>>> Sorry to bring up an old topic again but I can't find any 
>>> documentation of glycan refinement on the Phenix website and previous 
>>> threads don't seem to answer my question.
>>> 
>>> I am refining a structure with N-linked glycans.  I have inserted 
>>> several "apply_cif_link" records to my .eff file to define the glycan 
>>> topologies (see below) and fed in a cif file for BMA.  After fiddling 
>>> for quite some time now, I have managed to get around all the error 
>>> messages people have described previously to disappear.  Yet, 1) none 
>>> of my linkages appear to be enforced and 2) even the restraints on 
>>> the individual sugars appear to be unenforced (or at least, 
>>> insufficiently so).  For example, some rings are rendered almost 
>>> flat, or flipped into boat or otherwise distorted conformations, and 
>>> glycosidic bonds stretch to>2 A.
>>> 
>>> No obvious signs of trouble in the log file, running the latest 
>>> version (1.7, also tried 1.6.4), but my glycans look terrible 
>>> (despite some of the best density I have ever seen!  :-)  )
>>> 
>>> Any suggestions would be appreciated!
>>> 
>>> Thanks,
>>> 
>>> Damian Ekiert
>>> 
>>> 
>>> 
>>> From my .eff file:
>>> 
>>>     apply_cif_link {
>>>       data_link = NAG-ASN
>>>       residue_selection_1 = chain X and resname NAG and resid 401
>>>       residue_selection_2 = chain A and resname ASN and resid 91
>>>     }
>>>     apply_cif_link {
>>>       data_link = NAG-ASN
>>>       residue_selection_1 = chain Y and resname NAG and resid 401
>>>       residue_selection_2 = chain C and resname ASN and resid 91
>>>     }
>>>     apply_cif_link {
>>>       data_link = NAG-ASN
>>>       residue_selection_1 = chain Z and resname NAG and resid 401
>>>       residue_selection_2 = chain E and resname ASN and resid 91
>>>     }
>>>     apply_cif_link {
>>>       data_link = BETA1-4
>>>       residue_selection_1 = chain X and resname NAG and resid 402
>>>       residue_selection_2 = chain X and resname NAG and resid 401
>>>     }
>>>     apply_cif_link {
>>>       data_link = BETA1-4
>>>       residue_selection_1 = chain Y and resname NAG and resid 402
>>>       residue_selection_2 = chain Y and resname NAG and resid 401
>>>     }
>>>     apply_cif_link {
>>>       data_link = BETA1-4
>>>       residue_selection_1 = chain Z and resname NAG and resid 402
>>>       residue_selection_2 = chain Z and resname NAG and resid 401
>>>     }
>>>     apply_cif_link {
>>>       data_link = BETA1-4
>>>       residue_selection_1 = chain Y and resname BMA and resid 403
>>>       residue_selection_2 = chain Y and resname NAG and resid 402
>>>     }
>>>     apply_cif_link {
>>>       data_link = ALPHA1-3
>>>       residue_selection_1 = chain Y and resname MAN and resid 404
>>>       residue_selection_2 = chain Y and resname BMA and resid 403
>>>     }
>>> _______________________________________________
>>> phenixbb mailing list
>>> phenixbb at phenix-online.org
>>> http://phenix-online.org/mailman/listinfo/phenixbb
>> 
>> 
> 
> 
> -- 
> Engin Özkan
> Post-doctoral Scholar
> Laboratory of K. Christopher Garcia
> Howard Hughes Medical Institute
> Dept of Molecular and Cellular Physiology
> 279 Campus Drive, Beckman Center B173
> Stanford School of Medicine
> Stanford, CA 94305
> ph: (650)-498-7111
> 
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb




More information about the phenixbb mailing list