If you merge data with AIMLESS  I + and I - are always reported seperately
and it reports a CCanom checking whether two randomly selected data halves
have a correlation in the  anom signal .

Often *IF* you can position the S atoms the phasing steps work, but it can
be challenging to get the sites . We use SHELXD but it can be confused by
spurious "anomalous signal".   Phasing can more often bootstrap its way out
of trouble by various density modification and averaging tricks..
Eleanor




On 3 April 2018 at 16:16, Raghurama P Hegde <[email protected]> wrote:

> Hi Manoj,
>
>
>
> Like Eleanor has pointed out a very accurate data may get you a solution.
> If you have crystals that diffract well and give you data with high
> I/sigma(I) you can attempt to find the positions of the S atoms and use
> that to arrive at a structure. Using lysozyme as a test case we have shown
> that structure can be solved from a routine data set collected at 1 Å (with
> 10 S but a f” of ~0.28), with anomalous multiplicity ~12. Details are
> published in this paper (yes it’s a shameless plug for our paper! 😊):
>
>
>
> Hegde et al. (2017), The hidden treasure in your data: phasing with
> unexpected weak anomalous scatterers from routine data sets, *Acta
> Crystallogr F Struct Biol Commun,* *73*, 184-195.
>
> http://scripts.iucr.org/cgi-bin/paper?S2053230X17002680.
>
>
>
> However the data was not collected at a home source but at a synchrotron
> beamline. There could also be some weak anomalous scatterers acquired
> during crystallization, from the crystallization condition, so you could
> collect a data set to check for the presence of anomalous signal in it and
> attempt experimental phasing. That could give you pointers on a future
> course of action like going to a synchrotron beamline to collect stronger,
> more accurate data or prepare heavy atom derivates, like Eleanor has
> pointed out.
>
>
>
> As for data processing you’d have to process the data with Friedel mates
> pairs kept separate so that the anomalous differences can be used to find
> positions of the S atoms.
>
>
>
> Hope that helps,
>
> Raghu
>
>
>
> *From:* CCP4 bulletin board <[email protected]> *On Behalf Of *Eleanor
> Dodson
> *Sent:* Tuesday, April 3, 2018 20:17
> *To:* [email protected]
> *Subject:* Re: [ccp4bb] Sulphur SAD at home source
>
>
>
> Well - the S f" is only ~ 0.5 at Cu Kalpha so the signal will be very
> weak..
>
> Very accurate data may get a solution but you first have to position the S
> atoms...
>
> Much easier to try to make a heavy atom derivative!
>
> Eleanor
>
>
>
> On 3 April 2018 at 15:26, Manoj Saxena <00001d16aa30e8a1-dmarc-
> [email protected]> wrote:
>
> Hi All,
>
>
>
> I am writing to seek advice on doing  sulphur SAD data collection
>
> at Cu based home source for a protein that is 12 KDa and has 6 S atoms.
>
> I have seen some links online and some references but would be grateful if
>
> you can share your know-how for success with this.
>
> Like what multiplicity of data would be good to aim for and
>
> data processing tips.
>
> Inputs from people who have tried and failed would also be highly
> appreciated.
>
>
>
> Thank you
>
> Manoj Saxena
>
> University of Puerto Rico
>
>
>
>
>
>
>

Reply via email to