structure factors for PDB deposition
Hello all, based on ${prefix}_001_f_model.mtz from phenix.refine, I would like to prepare a cif file with structure factors. FOBS, SIGFOBS, R_FREE_FLAGS, FMODEL, PHIFMODEL, FOM are available in (+) and (-) versions, presumably representing Bijvoet pairs. My questions are: 1) are FMODEL, PHIFMODEL the most appropriate columns for calculated amplitudes and phases for the purpose of PDB deposition? 2) I understand that the PDB sf_tool does not accommodate separate (+)/(-) columns for some of the items. A job for phenix.reciprocal_space_arrays and a reflection file without anomalous differences? Alternatively, I imagine that phenix has a tailored procedure just for the preparation of PDB deposition, but I may not have used the correct web search terms to find the instructions. Thank you in anticipation, Wolfram Tempel
Hi Wolfram, each phenix.refine run creates a MTZ file that contains four sections of data: 1) original data as supplied at input; 2) data actually used in refinement (which may miss a few outliers, truncated by resolution and/or sigma if requested by the user, Fobs obtained from Iobs using French&Wilson method, etc); 3) Fmodel: total model structure factors including all scales, etc. They can be used to reproduce reported R-factors exactly; 4) Map coefficients: 2mFo-DFmodel (with and without filling missing Fobs), mFobs-DFmodel and anomalous difference map (f input data is anomalous: that is contains Fobs(+) and Fobs(-)). R-free flags are part of 1) and 2) above. That's the file that is supposed to go to PDB one way or another (as MTZ or CIF). Pavel On 1/22/13 9:28 AM, wtempel wrote:
Hello all, based on ${prefix}_001_f_model.mtz from phenix.refine, I would like to prepare a cif file with structure factors. FOBS, SIGFOBS, R_FREE_FLAGS, FMODEL, PHIFMODEL, FOM are available in (+) and (-) versions, presumably representing Bijvoet pairs. My questions are: 1) are FMODEL, PHIFMODEL the most appropriate columns for calculated amplitudes and phases for the purpose of PDB deposition? 2) I understand that the PDB sf_tool does not accommodate separate (+)/(-) columns for some of the items. A job for phenix.reciprocal_space_arrays and a reflection file without anomalous differences? Alternatively, I imagine that phenix has a tailored procedure just for the preparation of PDB deposition, but I may not have used the correct web search terms to find the instructions. Thank you in anticipation, Wolfram Tempel _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
On 22 January 2013 10:03, Pavel Afonine
Hi Wolfram,
each phenix.refine run creates a MTZ file that contains four sections of data:
1) original data as supplied at input; 2) data actually used in refinement (which may miss a few outliers, truncated by resolution and/or sigma if requested by the user, Fobs obtained from Iobs using French&Wilson method, etc); 3) Fmodel: total model structure factors including all scales, etc. They can be used to reproduce reported R-factors exactly; 4) Map coefficients: 2mFo-DFmodel (with and without filling missing Fobs), mFobs-DFmodel and anomalous difference map (f input data is anomalous: that is contains Fobs(+) and Fobs(-)).
R-free flags are part of 1) and 2) above.
That's the file that is supposed to go to PDB one way or another (as MTZ or CIF).
Pavel
On 1/22/13 9:28 AM, wtempel wrote:
Hello all, based on ${prefix}_001_f_model.mtz from phenix.refine, I would like to prepare a cif file with structure factors. FOBS, SIGFOBS, R_FREE_FLAGS, FMODEL, PHIFMODEL, FOM are available in (+) and (-) versions, presumably representing Bijvoet pairs. My questions are: 1) are FMODEL, PHIFMODEL the most appropriate columns for calculated amplitudes and phases for the purpose of PDB deposition? 2) I understand that the PDB sf_tool does not accommodate separate (+)/(-) columns for some of the items. A job for phenix.reciprocal_space_arrays and a reflection file without anomalous differences? Alternatively, I imagine that phenix has a tailored procedure just for the preparation of PDB deposition, but I may not have used the correct web search terms to find the instructions. Thank you in anticipation, Wolfram Tempel ______________________________**_________________ phenixbb mailing list [email protected] http://phenix-online.org/**mailman/listinfo/phenixbbhttp://phenix-online.org/mailman/listinfo/phenixbb
______________________________**_________________ phenixbb mailing list [email protected] http://phenix-online.org/**mailman/listinfo/phenixbbhttp://phenix-online.org/mailman/listinfo/phenixbb
Hi Wolfram, You can use the option write_reflection_cif_file=True in phenix.refine to output an mmCIF reflection file that contains the same data as in the MTZ file as described by Pavel. I recently fixed a bug regarding output of anomalous arrays to the mmCIF file, so it would probably be best to try with the latest nightly build The use of mmCIF for PDB deposition is an area we are actively working on right now, so we expect things to improve in this area in the near future. Cheers, Richard -- Richard Gildea Software Developer Physical Biosciences Division Lawrence Berkeley National Laboratory 1 Cyclotron Rd Mail Stop 64R0121 Berkeley CA 94720-8118
On Tue, Jan 22, 2013 at 11:07 AM, Richard Gildea
You can use the option write_reflection_cif_file=True in phenix.refine to output an mmCIF reflection file that contains the same data as in the MTZ file as described by Pavel. I recently fixed a bug regarding output of anomalous arrays to the mmCIF file, so it would probably be best to try with the latest nightly build
FYI: we haven't had a new nightly build available for more than a week due to unrelated bugs, so you're going to have at least another day for this fix. -Nat
participants (4)
-
Nathaniel Echols -
Pavel Afonine -
Richard Gildea -
wtempel