Print

Print
Second moments of Z but not twinning
Hi Folks,
 
Thanks for all your replies on this. It turns out that the two responses which were closest were Eleanor Dodson pointing me to the CCP4 documentation, where it says "you have probably made a mess of processing" and Peter Zwart saying that the second moment etc. are very sensitive to errors in the measurements.
 
So, it turns out that I was not properly rejecting the outliers that XDS CORRECT told me to reject, and there were quite a few of them (several hundred from 100,000 reflections) - removing these made the moment plots much more reasonable.
 
However, thanks to everyone who replied - lots of interesting cases to follow up on and things to check for in future.
 
Cheers,
 
Graeme

From: CCP4 bulletin board [mailto:[log in to unmask]] On Behalf Of Winter, G (Graeme)
Sent: 01 June 2007 11:04
To: [log in to unmask]
Subject: [ccp4bb] Second moments of Z but not twinning

Hi Folks,

I was wondering if there is a document somewhere explaining some of the more interesting things which can be interpreted from the plot of second moment of Z = I'/<I'> anywhere? I've looked around and I can't find much beyond the things pertaining to twinning, where it will have the magical values of 2.0 or 1.5. What about if it is say 2.0 around the low resolution regions but gets much bigger?

Looking at the truncate source code there is a note to the effect that the second moments may be greater than two in when you have pseudotranslational symmetry. In this case, however, I would expect that the effect would be uniform across the resolution ranges… any thoughts?

Any pointers would be greatly appreciated.

Thanks,

Graeme