does phenix.refine ignore F-obs, SIGF-obs?
Hi All, I have an mtz file that has: I-obs SIGI-obs F-obs SIGF-obs I want phenix.refine to use F-obs and SIGF-obs, but even though I provide : labels = "F-obs,SIGF-obs" I think phenix.refine is seeing the I-obs,SIGI-obs and recalculating the structure factors from those values instead of using the Fs that I provide in the mtz. Is this the case? Other than stripping the I-obs,SIGI-obs from the input mtz, is there a way to force phenix to use my structure factors? Thanks, Scott ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Scott Classen, Ph.D. SIBYLS Beamline 12.3.1 http://bl1231.als.lbl.gov Advanced Light Source Lawrence Berkeley National Laboratory 1 Cyclotron Rd MS6R2100 Berkeley, CA 94720 C) 510.206.4418 O) 510.495.2697 Beamline) 510.495.2134 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Scott, Try adding: refinement.input.xray_data.labels="F-obs,SIGF-obs" to the command line. Other alternatives would be to edit the .def file, or create a stripped down version of that using: phenix.refine mymodel-refine_002.def --diff-params > mystart.def and edit as necessary. Phil Jeffrey Princeton Scott Classen wrote:
Hi All,
I have an mtz file that has:
I-obs SIGI-obs F-obs SIGF-obs
I want phenix.refine to use F-obs and SIGF-obs, but even though I provide : labels = "F-obs,SIGF-obs"
I think phenix.refine is seeing the I-obs,SIGI-obs and recalculating the structure factors from those values instead of using the Fs that I provide in the mtz.
Is this the case? Other than stripping the I-obs,SIGI-obs from the input mtz, is there a way to force phenix to use my structure factors?
Thanks, Scott
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Scott Classen, Ph.D. SIBYLS Beamline 12.3.1 http://bl1231.als.lbl.gov Advanced Light Source Lawrence Berkeley National Laboratory 1 Cyclotron Rd MS6R2100 Berkeley, CA 94720 C) 510.206.4418 O) 510.495.2697 Beamline) 510.495.2134 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
Hi Scott,
I have an mtz file that has:
I-obs SIGI-obs F-obs SIGF-obs
I want phenix.refine to use F-obs and SIGF-obs, but even though I provide : labels = "F-obs,SIGF-obs"
by default phenix.refine takes I-obs,SIGI-obs. You need to tell phenix.refine explicitly that you want to use F-obs,SIGI-obs.
I think phenix.refine is seeing the I-obs,SIGI-obs and recalculating the structure factors from those values instead of using the Fs that I provide in the mtz.
Yes, this is what it does.
Other than stripping the I-obs,SIGI-obs from the input mtz, is there a way to force phenix to use my structure factors?
Yes, specify them explicitly: input.xray_data.labels="F-obs" Let me know if you have any problem or questions! Pavel.
Hello, On Mar 11, 2010, at 10:03 AM, Pavel Afonine wrote:
Hi Scott,
I have an mtz file that has:
I-obs SIGI-obs F-obs SIGF-obs
I want phenix.refine to use F-obs and SIGF-obs, but even though I provide : labels = "F-obs,SIGF-obs"
by default phenix.refine takes I-obs,SIGI-obs. You need to tell phenix.refine explicitly that you want to use F-obs,SIGI-obs.
OK it looks like phenix.refine IS using the F-obs and SIGF-obs, but it is still rejected a couple thousand reflections that have values <= 0. Can phenix.refine be directed to use F-obs <= 0? in particular can either of these flags be given slight negative numbers? input.xray_data.sigma_fobs_rejection_criterion= -0.1 input.xray_data.sigma_iobs_rejection_criterion= -0.1 BTW what does this flag do? input.xray_data.ignore_all_zeros= True Thanks, Scott
Hi Scott,
Can phenix.refine be directed to use F-obs <= 0?
No. See: http://www.phenix-online.org/version_docs/1.6.1-336/refinement.htm section "Remark on using amplitudes (Fobs) vs intensities (Iobs)".
in particular can either of these flags be given slight negative numbers? input.xray_data.sigma_fobs_rejection_criterion= -0.1 input.xray_data.sigma_iobs_rejection_criterion= -0.1
BTW what does this flag do?
It rejects reflections by sigma. In the past people used to cut the data at 2sigma or so. It is 0 by default in phenix.refine. Pavel.
OK, problem solved. Thanks to Peter Zwart. My "original" mtz file was generated from a CNS cv file many moons ago. I must have done this conversion in such a way that the F-obs were calculated from the input IOBS (some of which were slightly negative). This resulted in the mtz file indeed having ~3200 F-obs with values of zero. It was my intention during the original conversion process to keep the FOBS values from the cv file and merely import them into the mtz. This counts as user error. Thanks for all the quick responses and for the A+ user support. Scott On Mar 11, 2010, at 1:21 PM, Pavel Afonine wrote:
Hi Scott,
Can phenix.refine be directed to use F-obs <= 0?
No. See:
http://www.phenix-online.org/version_docs/1.6.1-336/refinement.htm
section "Remark on using amplitudes (Fobs) vs intensities (Iobs)".
in particular can either of these flags be given slight negative numbers? input.xray_data.sigma_fobs_rejection_criterion= -0.1 input.xray_data.sigma_iobs_rejection_criterion= -0.1
BTW what does this flag do?
It rejects reflections by sigma. In the past people used to cut the data at 2sigma or so. It is 0 by default in phenix.refine.
Pavel.
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
participants (3)
-
Pavel Afonine -
Phil Jeffrey -
Scott Classen