Dear Pu,
If I may add a few remarks those already made by George and Clemens, I
would say that you cannot expect two heavy atom solutions obtained in
separate runs of SHELXD (or any other substructure solution program) from
two distinct sets of differences to be consistent with each other in terms
of origin and enantiomorph - hence the need for checks by cross-difference
Fouriers or (much better) log-likelihood gradient maps for each set of
differences phased by the other. This could get even worse in the case of
polar space groups (e.g. P2(1) ) where a coordinate shift along the polar
axis is left undetermined in each substructure solution. This is a very old
problem - see for instance the method of Bragg ellipses.
One way of avoiding this in your case would have been to treat your
isomorphous and anomalous differences jointly by trying to get estimates for
FA values, as if you had a 2-wavelength experiment. This was done under the
old terminology via the Matthews "combined difference Patterson". Then,
solving for your substructure from that combined Patterson would have
produced an automatically consistent solution for both the isomorphous and
the anomalous differences.
George: wouldn't XPREP have done that if it had been given the SIRAS
data all at once, instead of separately handling these as SAD and SIR data?
With best wishes,
Gerard.
--
On Fri, Aug 20, 2010 at 04:32:37PM +0100, Pu Gao wrote:
> Dear all,
>
> My case seems to have two factors: 1. origin choice. and 2.
> handedness/enantiomorph.
>
> And this kind of "problem" seems to be very common during the experimental
> phasing process. And I learned a lot informations both theoretical and
> practical from your replies.
>
> Many thanks for all the replies.
>
> Pu
--
===============================================================
* *
* Gerard Bricogne g...@globalphasing.com *
* *
* Global Phasing Ltd. *
* Sheraton House, Castle Park Tel: +44-(0)1223-353033 *
* Cambridge CB3 0AX, UK Fax: +44-(0)1223-366889 *
* *
===============================================================
