[sharp-discuss] Heavy Atom Negative Density

[Clemens Vonrhein]

Ed,

Although SHARP has the notion of an 'unperurbed native' structure
factor, it will by default take care of cases like SAD or MAD - where
the heavy atom substructure is part of the data.

It's a bit more complicated in cases like MAD+native or MIR (with a
derivative as reference). But in all 'classical' cases the resulting
electron density map should show exactly what you expect, ie a
'native' for MIR/SIR and the protein plus any scatteres for SAD/MAD.

And you're right: even if SHARP put out centroid structire factors for
the 'native' only in this MAD case, there should be zero density (and
not -5 sigma).

I hope that the hierachy is right for that MAD dataset:

 C-1                 Compound
   X-1               Crystal
     W-1             Wavelenght
       B-1           Batch
     W-2
       B-1
     W-3
       B-1

So the occupancies/B (at the crystal level) are all identical. The
only difference is (apart from non-isomorphism) in the f'/f'' values
used at the batch level. Therefore, a thorough inspection of the ANO
residual maps at each wavelength might help.

Cheers

Clemens

On Tue, Jan 29, 2002 at 04:08:11PM -0800, Edward A. Berry wrote:
> Just a thought- could sharp be generating Fo/Phic coefficients for an 
> artificial "native" protein lacking the irons? I think it would do, 
> if the iron sites were used as "heavy atoms" in phasing, and they are 
> listed as sites in every compound/crystal. Still, I would only expect
> zero density in that case, not minus 5 sigma, unless the occupancies 
> in the reference crystal were badly overestimated.
> Ed

-- 

***************************************************************
* Clemens Vonrhein, Ph.D.          vonrhein@GlobalPhasing.com
*
*  Global Phasing Ltd.
*  Sheraton House, Castle Park     Tel: +44-(0)1223-353033
*  Cambridge CB3 0AX, UK           Fax: +44-(0)1223-366889
*--------------------------------------------------------------
* BUSTER Development Group      (http://www.globalphasing.com)
***************************************************************