[sharp-discuss] Ta6Br12 cluster with run_autoSHARP

[Clemens Vonrhein]

Hi Francis,

On Wed, Aug 05, 2015 at 03:16:00PM -0400, Francis E Reyes wrote:
> Hi all, 
> 
> I'm curious about how to look for Ta6Br12 clusters using the run_autoSHARP.sh utility
> 
> I supplied the line
> 
> '-ha "Ta6Br12:Ta" -nsit 10 -wvl 0.97921 hrem -mtz TaBr.mtz' but it
> seems to be only filling in Ta sites via the SPHCLUSTER Ta6Br12:Ta
> and not including Ta6Br12:Br.

Correct: you only asked for "Ta6Br12:Ta".

> I have a (semi) isomorphous native and therefore, inclusion of both
> scattering factors makes sense.

Yes - but although there is functionality within autoSHARP to specify
different types of heavy atoms, there is currently no feature that
would constrain different atomic subtypes (of such a cluster) to the
same position (G-SITE) and occupancy/B-factor (HAT_OCC/HAT_B).

However, that functionality is available when running SHARP directly,
probably based on your initial autoSHARP run with the "Ta6Br12:Ta"
alone. Of course, the (correct) idea of specifying both Ta and Br
atoms and scattering informaiton for the spherical-averaged cluster
feature assumes (a) perfectly located and (b) perfectly disordered
TaBr clusters as well as (c) HA signal to high enough resolution that
the inner Ta- and outer Br-shell are distinguishable.

Even if that is met, this is not really a project for autoSHARP
then. You could use autoSHARP for finding the TaBr positions using the
Ta atoms of a SPHCLUSTER. And then _if_ there is indication of the Br
atoms (ANO or ISO residual maps ... or maybe a too large HAT_B), you
could then switch to SHARP. This then gives complete control to ensure
the C-SITEs use the same G_SITEs and all that.

Does that make sense?

Cheers

Clemens Vonrhein & Claus Flensburg