Re: [ccp4bb]: Re: Question about I to F conversion

["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         ***

> 
> Maybe I should restate my question:
> 
> I have a file of intensities, processed with a -3 sigma cut to preserve
> the negative reflections.  If I convert this to amplitudes without any
> Wilson scaling and then apply a 2 sigma cut I get 89% completeness.  If
> I take the same starting file of intensities, convert it to amplitudes
> with Wilson scaling and then apply a 2 sigma cut I get 97%
> completeness.


 Wilson scaling does not change anything - it only applies the same
scale/TF to both I and Sigma.

 When you convert Is to Fs the SigF become roughly SigF = SigI/2F ;
see the TRUNCATE documentation for the exact formula..
After this of course the % of Fs< 2SigF will be different to the % of I
< 2SigI.

There is also the problem of what to do with the physically meaningless
I < 0.
TRUNCATE has a sensible option to convert such I to 0 , (which will
yield F 0 too), and to modify the weak intensities to fit the predicted
intensity statistics. 

> 
> My problem is that I don't understand how the Wilson scaling works with
> respect to the sigma values.  If a particular reflection is scaled up,
> how is the corresponding sigma changed?  I'm sorry to bother everyone, I
> could probably answer this question myself if I had the French and
> Wilson paper, but I don't have easy access to the older Acta. journals
> right now.


 See above - nothing to do with Wilson scaling...


Another Q - why are you doing a 2 Sigma cutoff for anything. Maximum
likelihood refinement requires that you use all the data, including the
weak..

 Eleanor