Re: [ccp4bb]: unmeasured data in map-calculations

[Jan Abendroth <Jan.Abendroth@uni-koeln.de>]

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

Dear Ed,
thanks for the reply!

However I simply still don't understand one phenomenon:
- the 2fofc-maps in both cases (with and without unobserved reflexes set
to Fc) are extremely similar. They differ just in very tiny details.
- Significant differences are in the fofc-maps. 
  But why does it make a difference between
	a) disregarding unobserved reflexes ie. setting them to zero or
	b) setting Fo=Fc which makes them 0 in a FoFc map?

Kind regards
Jan

Edward Berry wrote:
> 
> Jan Abendroth wrote:
> > My data set is quite incomplete in the highest resolution shell. When I
> > calculate maps with the cns script model_map the sigmaA-weighted
> > fofc-map (mFoDFc) has many "unreasonable" peaks at 3sigma-level. When I
> > make cns to take Fc for the unmeasured Fo's the sigmaA-weighted fofc-map
> > shows much fewer and much more reasonable peaks, simply this map becomes
> > significanly better.
> >
> > This seems to make sense, but what is the physical/mathematical reason
> > for this?
> > How treats cns unmeasured hkl's in the "normal" mode in difference maps?
> > Is there a related ccp4-option in refmac/fft?
> >
> 
> My 2 cents worth, and perhaps the experts will also forgive me in the spirit
> of the season:
> 
> fft has keyword fillin which causes it to substitute Fc for Fo in nFo-mFc maps.
> 
> If this is not done, any map calculation program has little option but to
> omit the unobserved reflections, which is equivalent to setting them to zero,
> which I think might account for your inexplicable peaks.
> 
> The philosophy of fillin is that the Fc is a better estimate of the
> unobserved reflection than is zero, so will give a better map. But beware
> of model bias! You will get the most beautiful map, with absolutely no
> peaks that don't make sense, when you have zero completeness and use fillin:
> 2Fc-Fc = Fc so you are making an Fc/PHIc map. The Fourier transform
> is reversible and you will get back exactly the density of your model.
> 
> If you are doing density modification it is very important to use fillin
> according to I. Rayment- Otherwise you are solving for the best set of
> phases consistent with the flat solvent, the NCS, AND A VALUE OF ZERO
> FOR ALL THE MISSING REFLECTIONS. If you use fillin the value of missing F's
> gets dragged along by neighboring reflections in reciprocal space, just as
> the phases of all reflections do, and when you go back and measure the
> missing reflections they often come out close to the value of Fc from DM
> with fillin (again according to Rayment).
> 
> Ed

-- 
------------------------------
Jan Abendroth
Institut fuer Biochemie
Universitaet Koeln
Zuelpicher Strasse 47
D-50674 Koeln

Tel: +49-(0)221 470 6455 (Labor)
                    6445 (Buero)
Fax: +49-(0)221 470 5092