Hi,
Some
background waffle: XDS has a program called "cellparm" for calculating a
weighted average of the unit cell parameters over the whole data set, and indeed
over multiple sets. I would guess that denzo has something similar. Using Mosflm
you would usually prerefine the unit cell then fix this during
integration.
The
unit cell parameters are varying to make the predictions of the spots fit best
to the observations, and usually the variation is mopping up some hidden changes
which are not explicitly being accounted for. Getting a weighted average is
definately the way forward (I expect scalepack can do
this).
Cheers,
Graeme
-----Original Message-----Hi all,
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]On Behalf Of Alessio Accardi
Sent: 09 January 2006 21:06
To: [email protected]
Subject: [ccp4bb]: Unit Cell Parameters
I am very sorry if I have a somewhat trivial question, but I am quite new to the crystallography world.
I process my data with Denzo followed by Scalepack and I noticed that the unit cell parameters in the scalepack .log file and in the final output *.sca file are different; not by much, at most around 1-1.5%.
This is the readout from the middle of the scalepack.log file:
Hkl's refined: 379123 N.Chi**2: 1.227 Decrease: 0.000 * 379123 = 0.5
Film # a b c alpha beta gamma crysz crysy crysx mosaicity
1 232.332 97.983 170.963 90.000 131.210 90.000 75.087 18.917 135.045 1.082
2 232.332 97.983 170.963 90.000 131.210 90.000 75.126 18.942 135.056 1.082
3 232.332 97.983 170.963 90.000 131.210 90.000 75.124 18.933 135.057 1.082
4 232.332 97.983 170.963 90.000 131.210 90.000 75.126 18.930 135.056 1.082
5 232.332 97.983 170.963 90.000 131.210 90.000 75.117 18.908 135.052 1.082
And this is the readout from the top of the output.sca file:
230.822 97.684 170.202 90.000 131.114 90.000 c2
First of all which of the two sets of values is the correct one to use in refinement? I would guess that the right one is the one in the .sca file as the hkl's and intensities refer to that unit cell, but I am not sure. Also, why are the values different? do they reflect different stages of refinement of the processed data? if so, is the one in the .sca file the more refined one?
Thank you
Alessio
Brandeis University
