Re: [ccp4bb]: French and Wilson scaling

["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         ***

Vidyasankar Sundaresan wrote:
> 
> ***  For details on how to be removed from this list visit the  ***
> ***          CCP4 home page http://www.ccp4.ac.uk         ***
> 
> I have data processed and scaled in d*trek, which I'm trying to convert to
> MTZ, using the "Import Scaled Data" feature in CCP4. The maximum resolution in
> my data is 3.5 Angstroms, and I seem to have some trouble with the French and
> Wilson scaling protocol. Under the option for excluding reflections from
> Wilson scaling, I've tried leaving it unchecked and also setting a resolution
> range. Either way, I get a scale factor value in the thousands. Is this
> acceptable? Or should I just use the option to derive structure factors from
> square root of intensities instead?
> 
> Vidyasankar


 The value of the scale factor will be governed roughly by the number of
atoms in the asymmetric unit. Have you set this correctly? 
I am not familiar enough with d*trek output values - but most proteins
should give <F> values betwen 100 and 500. Is this true for your output?

 The other problem may be due the the data resolution - look at the
plots and see if the distribution looks "normal". If it falls off very
quickly, with 3.5A data you may have a poor estimate of the B factor and
therefore a weird value of the scale.. It isnt applied to the data so as
long as the distribution is reasonable you neednt worry I dont think..

Do I need to expand on what is normality?

 Eleanor

> 
> MB 113, Stout lab
> Department of Molecular Biology
> The Scripps Research Institute
> 10550 North Torrey Pines Road
> La Jolla CA 92037
> 
> Tel: 858 784 8751 or 858 784 2426
> Fax: 858 784 2857