[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[ccp4bb]: Postrefinement in MOSFLM



***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***



Dear Rohini,

            In order to use more than 30 images you would need to edit
the mosflm source code to change the parameter NIMAX.

In the code, this parameter occurs many times in lines like:

PARAMETER (NIMAX=30)

You need to do a GLOBAL edit of the source files, to change ALL occurences of
NIMAX=30 
to 
NIMAX=60  (say).

You need to do this to all the ".f" and ".for" files in the directory
"mosflm" that is created when you have un-tared the mosflm tar file
from our ftp site.

If you are using the code distributed with CCP4, these files will be
in a directory similar to:

ccp4-4.2.2/ccp4-4.2.2-alf/x-windows/Mosflm/mosflm

(The exact directory name depends on how CCP4 has been installed).

For version 6.2.2 this includes the files:
mosflm_all_ip_inc.for, unoptimizable.f and control.f 

and for 6.2.0 also the file torefix.f

Then build a new executable.

HOWEVER
=======

I'm not sure why you want to use so many images for postrefinement.
The recommended way to refine cell parameters is to use two (or more)
segments of data widely separated in phi (ideally 90 degrees apart).
For monoclinic or triclinic spacegroups, use 3 segments at (say)
phi=0,45 and 90.

There is normally nothing to be gained from using a total of more than
10-15 images in the refinement (it just takes longer). The only
possible exception is if you have a very small unit cell, so there are
not many spots on each image, or if you have a very small oscillation
angle compared to the mosaic spread. Even if these cases, 30 images is
usually enough.

During integration, it is best to keep the cell parameters fixed, and
only use the post-refinement to refine the crystal orientation and
mosaic spread.

Best wishes,

 Andrew Leslie