[ccp4bb]: refining cell at low resolution

["Leslie A." <andrew@mrc-lmb.cam.ac.uk>]

***  For details on how to be removed from this list visit the  ***
***    CCP4 home page http://www.dl.ac.uk/CCP/CCP4/main.html    ***



It should be realised that there are two options for refining the cell
parameters in MOSFLM.

First, cell parameters are refined as an integral part of the
autoindexing. In order to obtain well determined values for all cell
parameters, spots from several images widely separated in phi should
be used, particularly for lower symmetry spacegroups. However, this
refinement does assume the same crystal orientation for all images.
Also, particularly at low resolution, there will be a high correlation
between cell lengths and the crystal to detector distance, so that any
error in the detector distance will result in an error in the refined
cell.


Second, the cell parameters can be refined using the Post Refinement
procedure (Refine Cell menu option or POSTREF keyword). Again, data
from more than one segment of reciprocal space should be used for best
results.

In practice, the best cell parameters are generally obtained by
post-refinement. However, for low resolution data (say les than 4A),
post-refinement will not generally give a very accurate cell, and the
parameters obtained from the autoindexing are more reliable. This will
almost certainly be true for data at 10A resolution. 


          Andrew Leslie