[phenixbb] How the sharpening were performed in PHENIX.

[Pavel Afonine]

Hi Hailiang,

> I was trying to do a B-factor sharpening using phenix GUI CreatMaps
> utilities. The output for the sharpened maps contains 2FOFCWT and
> PHI2FOFCWT columes, so I think they are sigmaA weighted 2Fo-Fc maps.

there is no need to guess. If you look at the "Map type" filed it will 
tell you which map you computed. Looking right now: "Map type" says 
"2mFo-DFc", and the corresponding labels are 2FOFCWT, PH2FOFCWT.
So the answer: yes, what you get is the maximum-likelihood weighted map.

> But
> from Brunger's paper, the sharpening coefficients seems directly applied
> on Fo, and I wonder what happened before we get the 2mFoDFc maps. Or, the
> sharpening coefficients have been applied to the 2mFoDFc maps?

I do not remember what's in that paper, my vague recollection is that 
the exp(B*s**2) is applied to F=map_coefficients. Anyway, in PHENIX the 
exp B-sharpening applied to the map coefficients (not to Fobs). You have 
two options here: provide your own B-factor to be used for sharpening or 
let PHENIX determine it automatically.

Pavel.