Hello all-
I am trying to merge a few partial Se datasets. I am certain that I will have to reindex one or more to make sure that + and - reflections are not jumbled up. Can someone confirm that xtriage take this into account in suggesting to reindex one dataset in relation to a reference dataset? ie. does it use something like maximizing the anomalous signal in suggesting whether or not reindexing is necessary? I think the below suggests as much but I wanted confirmation:
Automatic re-indexing
For each data set supplied (other than the first data set given), all possible re-indexing matrices are derived as described by Grosse-Kunstleve et al. (2005). Each data set isautomatically re-indexed using the matrix that maximizes the correlation of amplitudes
Have a good weekend.
-Todd
My Example:
Reference analyses
The following reindexing operators have been found:
-------------------------------------------------
| Operator | Correlation | matches (%) | choice |
-------------------------------------------------
| -k,h,l | 0.874 | 70.1 | <--- |
| h,k,l | 0.032 | 76.7 | |
-------------------------------------------------
If the data is reindexed with operator (-k,h,l), the correlation of
the intensities to the reference data is maximized.
sx205b_anom.sca
This can be done for instance with:
phenix.reflection_file_converter sx205b_anom.sca --change_of_basis="-k,h,l"