Hi Yarrow,

thanks for sending log files.
Completeness reported in log files reflects the content of input MTZ data files:

32.log:
(...)
F-obs:
(...)
Miller array info: (...)/reflectionsFreeR_1.38A.mtz:F_New(+),SIGF_New(+),F_New(-),SIGF_New(-),systematic_absences_eliminated
Observation type: xray.amplitude
Type of data: double, size=325598
Type of sigmas: double, size=325598
Number of Miller indices: 325598
Anomalous flag: True
Unit cell: (64.3967, 68.0966, 103.663, 90, 96.1167, 90)
Space group: P 1 21 1 (No. 4)
Systematic absences: 0
Centric reflections: 4910
Resolution range: 43.7941 1.3786
Completeness in resolution range: 0.901282
Completeness with d_max=infinity: 0.901259
(...)

and

35.log:
(...)
I-obs:
(...)
Miller array info: (...)/reflectionsFandIFreeR_1.38A.mtz:IMEAN_New,SIGIMEAN_New,systematic_absences_eliminated
Observation type: xray.intensity
Type of data: double, size=181940
Type of sigmas: double, size=181940
Number of Miller indices: 181940
Anomalous flag: False
Unit cell: (64.3967, 68.0966, 103.663, 90, 96.1167, 90)
Space group: P 1 21 1 (No. 4)
Systematic absences: 0
Centric reflections: 4910
Resolution range: 43.7941 1.3786
Completeness in resolution range: 0.992202
Completeness with d_max=infinity: 0.992164
(...)

The numbers above are reported before anything else is done, that is they reflect exactly what's in input MTZ files.

Somehow you managed to loose ~10% of data when converting Iobs into Fobs. How did you do this conversion? In theory this is possible since French&Wilson Iobs->Fobs conversion may result on some number of reflections lost. Note, when phenix.refine does IMEAN->Fobs conversion (35.log file) this does not result in ~10% loss.

So based on looking at log files I can say it's not phenix.refine that manipulated your data or reported it wrong. You lost reflections at previous steps.

Pavel

On 12/4/12 4:57 PM, Yarrow Madrona wrote:
Hi,

Here are the log files for refinement you asked for.

35.log is refined with Intensities. The MTZ file has a slightly different
name but it is the same MTZ. One had Intensities taken out of it because I
didn't think I needed them any longer.

32.log is refined with structure factors. Intensities were removed.

Thanks for your help!

-Yarrow