Dear Phil,

I wouldn't be very optimistic about this. I recently tried to solve a 50:50 twin structure with simulated data (prepared with the help of a program and data by James Holton), and failed. When I reduced the twinning fraction at some point the structure solution worked out well (as expected), but I didn't try to find the exact twin fraction for this to be possible.

It may nevertheless be possible, if the degree of HA substitution is high and the data are really good. However, I'd certainly worry about adding the complex structure factors of the heavy atom contributions which the phasing programs perform, because in this case they should rather add the intensities if the heavy atom structure factors are from twin-related sites (I haven't thought about this carefully ...), and only add the complex structure factors if they belong to the same twin domain. AFAIK none of the programs supports this, so the only way, it seems, would be to identify twin-related sites and leave one of them out of the phasing, thus trying to "phase only half of the observed intensity".

good luck,

Kay

Am 20:59, schrieb Phil Evans:
We're trying to solve a 50% twin crystal by experimental phasing (SAD&  SIRAS) 
since molecular replacement hasn't worked. The crystals diffract very well in 
apparent space group P62 2 2. A Hg derivative shows a strong apparent anomalous 
signal (from half-dataset correlations etc), but we've failed to solve it so far, 
nor even to decide the true space group.

Has anyone out there successfully solved a 50% twin by experimental phasing? 
Any suggestions or tips? It would be nice to know if we are not attempting the 
impossible

Phil Evans&  Rich Suckling



Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

Reply via email to