SOLVE automated phase determination
Web site
Data preparation
The recommended diffraction data format for input to SOLVE when processed
with DENZO/SCALEPACK is the "no merge original index" option in SCALEPACK.
"Anomalous" must be also specified in SCALEPACK to separate Friedel mates
in output. If "scale anomalous" is also specified, Friedel mates
will also be separate in calculating scaling factors and statistics (resulting
in lower R factors). The suggestion is *not* to use "scale anomalous"
unless observations for each reflection are highly redundant (e.g., high
symmetry space group or many images collected). I gather that it
is safe and conservative to *not* use "scale anomalous" in general, since
improvements in scaling factors when keeping Friedel mates separate are
probably minimal.
It will also be necessary to generate a scaled/merged dataset, at least
for a reference wavelength (this scaled/merged dataset will be required
for model building and refinement with CNS or w/ARP). The two example
SCALEPACK scripts below were used to generate the "no merge original index"
output and the scaled/merged output.
Note that he current version of SOLVE has limits on the amount of data
it can handle. For my P2(1) crystal, it could process data out to
2.0 A, but not to 1.7 A.
Setup
Command file
Output summary file