Difference between revisions of "Cppxfel Final Merge"

From cctbx_xfel
Jump to: navigation, search
(Created page with "The final merge in cppxfel is carried out in order to choose the final rejection criteria and recalculate the sigmas so that they better reflect the precision of i...")
 
 
Line 24: Line 24:
 
This uses the file allMerge5.mtz calculated using in the previous stage, and the merge-orientations.dat list which refer to the refined MTZ files (<code>ref-img*.mtz</code>) generated by post-refinement. This will produce a final merge and associated merging statistics.
 
This uses the file allMerge5.mtz calculated using in the previous stage, and the merge-orientations.dat list which refer to the refined MTZ files (<code>ref-img*.mtz</code>) generated by post-refinement. This will produce a final merge and associated merging statistics.
  
If <code>MERGE_ANOMALOUS</code> is set to <code>ON</code> then an anomalous data set will be generated instead of a normal merge.
+
If <code>MERGE_ANOMALOUS</code> is set to <code>ON</code> then an anomalous data set will be generated instead of a normal merge. This creates I(+), SIGI(+), I(-) and SIGI(-) columns containing the anomalously merged intensities.
  
 
== Output of merge.txt ==
 
== Output of merge.txt ==
  
In the output of merge.txt the correlations in each indexing hand are displayed for each image. The log file ends with new merging statistics. The correlation coefficient can be heavily altered due to the new recalculated sigma values and may appear less high than one would like. If in doubt, plot the intensities of the half-data sets against each other as explained in [[Cppxfel Statistics|cppxfel statistics]] to confirm their successful correlation.
+
In the output of merge.txt the correlations in each indexing hand are displayed for each image. The log file ends with new merging statistics. The correlation coefficient can be heavily altered due to the new recalculated sigma values and may appear less high than one would like. If in doubt, [[Cppxfel Statistics#Correlation between two images|plot the intensities]] of the half-data sets against each other to confirm their successful correlation.
  
 
The two halves of the data set are output to <code>half1Merge.mtz</code> and <code>half2Merge.mtz</code>, and the final merged data set will be output to <code>remerged.mtz</code>, which should be used in further structure solution. If the merge was set to merge anomalous information, then the appropriate file will be saved as <code>anomalous_diff.mtz</code>.
 
The two halves of the data set are output to <code>half1Merge.mtz</code> and <code>half2Merge.mtz</code>, and the final merged data set will be output to <code>remerged.mtz</code>, which should be used in further structure solution. If the merge was set to merge anomalous information, then the appropriate file will be saved as <code>anomalous_diff.mtz</code>.

Latest revision as of 17:25, 25 November 2015

The final merge in cppxfel is carried out in order to choose the final rejection criteria and recalculate the sigmas so that they better reflect the precision of individual reflections. This is carried out after post-refinement and is the final stage in generating a suitable MTZ file.

Contents of merge.txt

The contents of merge.txt from cppxfel.input_gen can be displayed.

cat merge.txt
ORIENTATION_MATRIX_LIST merge-orientations.dat
MATRIX_LIST_VERSION 2.0
INITIAL_MTZ allMerge5.mtz

OUTLIER_REJECTION_SIGMA 1.6
MERGE_ANOMALOUS OFF
RECALCULATE_SIGMA ON
CORRELATION_THRESHOLD 0.9

COMMANDS

MERGE

This uses the file allMerge5.mtz calculated using in the previous stage, and the merge-orientations.dat list which refer to the refined MTZ files (ref-img*.mtz) generated by post-refinement. This will produce a final merge and associated merging statistics.

If MERGE_ANOMALOUS is set to ON then an anomalous data set will be generated instead of a normal merge. This creates I(+), SIGI(+), I(-) and SIGI(-) columns containing the anomalously merged intensities.

Output of merge.txt

In the output of merge.txt the correlations in each indexing hand are displayed for each image. The log file ends with new merging statistics. The correlation coefficient can be heavily altered due to the new recalculated sigma values and may appear less high than one would like. If in doubt, plot the intensities of the half-data sets against each other to confirm their successful correlation.

The two halves of the data set are output to half1Merge.mtz and half2Merge.mtz, and the final merged data set will be output to remerged.mtz, which should be used in further structure solution. If the merge was set to merge anomalous information, then the appropriate file will be saved as anomalous_diff.mtz.