Re: [ccp4bb]: Anomalous differences in ccp4i

["Eleanor J. Dodson" <ccp4@ysbl.york.ac.uk>]

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

Phil Evans wrote:
> 
> ***  For details on how to be removed from this list visit the  ***
> ***          CCP4 home page http://www.ccp4.ac.uk         ***
> 
> Beware of calculating anomalous difference Pattersons in ccp4i. Don't
> use the option "anomalous difference Patterson". This calculates
> differences from F(+) and F(-), and nowhere are these set equal for
> centric data. Thus centric data get junk values and the Pattersons are
> rubbish
> 
> The option "Patterson using anom diff (D) data" is correct since DANOM
> is set to zero in Truncate for centrics.
> 
> I think this is probably a Truncate bug
> 
> The same may apply for setting up data for Rantan etc, but I haven't
> checked
> 
>         Phil


 Phil is I think overstating the case.
 The Pattersons should not be rubbish.  Any errors on the centric are
just the outward and visible form of the degree of error in ALL your
data. If they diverge seriously you have a problem which should be
addressed at the data processing stage.

 When calculating any sort of difference patterson, it is very important
to use the EXCLUDE cards sensibly, and to keep out rubbish terms. Have
centric terms not set to 0.00 will add a small noise component, but most
Pattersons should high peaks on top of a noise film quite happily. The
reason F+ is not set to F- for centric terms is to allow you to get some
feel for the degree of error in your data. It can only happen for those
who use MOSFLM/SCALA to process data; SCALEPACK and XDS force equality
on centric terms..

 Eleanor