[ccp4bb]: ccp4bb: Can a correct partial model degrade experimental phases?

[<wgscott@chemistry.ucsc.edu>]

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

Hi folks:

I was just asked if a correct partial model might degenerate the 
quality of phases when performing partial model phase combination, 
using sigma-a weighting.

Say I have an MIR or MAD phase set and build only 5 to 10% of the atoms 
into electron density.  Let's further assume that the model is simply 
incomplete, not
incorrect (in the sense of wrongly-placed atoms).

My initial answer was that it could only help or do nothing, but was 
then asked if omitting atoms from the molecule is the equivalent of 
having an "incorrect" model in the sense that most of the scattering 
atoms are being modeled in as nothing (or solvent).  Would a partial 
model having only 5 or 10% of the atoms therefore degrade the phases?  
If so, at what point would you break even?


Bill Scott