[phenixbb] phenix.refine with simulated annealing

Pavel Afonine pafonine at lbl.gov
Fri Dec 17 16:56:34 PST 2010


  Hi Hailiang,

> Hi there,
>
> I was trying to perform a SA refinement for PDB# 2VZ8 using the following
> phenix script:
>
> random_seed=2679941
> phenix.refine 2VZ8.pdb 2vz8.mtz simulated_annealing=true
> main.random_seed=${random_seed} output.prefix=${random_seed}
>> ${PDB}-phx-sa.log
> where 2VZ8.pdb was downloaded from rcsb, and 2vz8.mtz was converted from
> the cif file by ccp4-cif2mtz.

for future reference: you can use phenix.cif_as_mtz to convert a CIF 
data file to MTZ file.

> But it ended up with:
>
> *******************
> =============================== refinement start
> ==============================
>
> Traceback (most recent call last):
>    File
> "/users/hz9/TOOLS/PHENIX/phenix-1.6.4-486/phenix/phenix/command_line/refine.py",
> line 11, in<module>
>      command_line.run(command_name="phenix.refine", args=sys.argv[1:])
>    File
> "/users/hz9/TOOLS/PHENIX/phenix-1.6.4-486/phenix/phenix/refinement/command_line.py",
> line 89, in run
>      call_back_handler=call_back_handler)
>    File
> "/users/hz9/TOOLS/PHENIX/phenix-1.6.4-486/phenix/phenix/refinement/driver.py",
> line 1182, in run
>      call_back_handler       = call_back_handler)
>    File
> "/users/hz9/TOOLS/PHENIX/phenix-1.6.4-486/phenix/phenix/refinement/strategies.py",
> line 229, in refinement_machine
>      xray_structure = model.xray_structure,)
>    File
> "/users/hz9/TOOLS/PHENIX/phenix-1.6.4-486/phenix/phenix/refinement/strategies.py",
> line 133, in extract_from_pdb_header
>      xray_structure = xray_structure)
>    File
> "/users/hz9/TOOLS/PHENIX/phenix-1.6.4-486/cctbx_project/mmtbx/f_model.py",
> line 302, in __init__
>      twin_law_xyz = sgtbx.rt_mx(symbol=self.twin_law, r_den=12, t_den=144)
> RuntimeError: cctbx Error: Parse error: unexpected character:
>    NULL
> ************************

This is a known bug that will be fixed in a newer version. A work-around 
for you is: locate and remove these lines from PDB file header:

REMARK 3 TWINNING INFORMATION.
REMARK 3 FRACTION: NULL
REMARK 3 OPERATOR: NULL

This structure (2VZ8) was refined using phenix.refine, and, believe me 
or not, phenix.refine would not put these lines if ML refinement target 
was used (which is the case for this model). Therefore I'm puzzled about 
how the above lines ended up in this PDB file. This reminds me to repeat 
once again: any post-refinement manipulation of a PDB file with refined 
mode is bad.

Pavel.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20101217/dc2610d4/attachment-0002.htm>


More information about the phenixbb mailing list