[phenixbb] Custom ligand - confirming the correct scattering definition

[George Phillips]

Pavel,
I am following up on this question of the oxidation state of metals. It is important to deal with this properly to get the most out of the data and to able to show difference maps in papers that make sense.
Where are the scattering factors stored (it1992's for example)?  One could hack an oxidation state in for the standard numbers to get it right (ugly but effective). Alternatively, maybe the scattering factor tables and element id's could be expanded?  Is this all hard coded or are there libraries or tables that are parsed?
THanks for any comments, as always.
George N. Phillips, Jr., Ph.D.
Professor of Biochemistry and of
     Computer Sciences
University of Wisconsin-Madison
433 Babcock Dr. Madison, Wi 53706
Phone/FAX (608) 263-6140

Hi Andy,

currently phenix.refine will not make difference between say "Fe" and 
"Fe+2". So if you see some artifacts at mFo-DFc map around your Fe, that 
might be explained by not taking "+2" into account. Refining occupancy 
of Fe would fix the map appearance (most likely), but I would not do 
that, since it will rather hide the problem than properly address it 
(and of course it  would be misleading too).

Regarding "n_gaussian table": it is just a dynamic approximation to the 
standard scattering factor table (mostly done to optimize runtime; the 
accuracy is not compromised). For more details see:

Grosse-Kunstleve RW, Sauter NK, Adams PD:
cctbx news
Newsletter of the IUCr Commission on Crystallographic Computing 2004, 
3:22-31.

Pavel.


On 11/25/09 12:37 PM, Andy Torelli wrote:
> To the Phenix group:
>
> 	I've created a .cif file for a ligand that includes metal atoms.   The 
> metal atoms are defined in the .cif file with their element name, but in 
> actuality, they should be defined according to their oxidation state so 
> that the correct scattering parameters are used during refinement.
>
> 	In CNS, this meant choosing an "atom type" that would be parsed to 
> match the correct scattering definition listed in the scatter.lib file. 
>   I see that phenix uses the n_gaussian table for scattering values. 
> Where can I view this and the other scattering table options?  Like CNS, 
> do I just need to use the name of the element with the correct 
> oxidation state from the table (e.g. "Fe+2") in my .cif file to ensure 
> that the correct scattering definition is used during refinement and map 
> calculation?
>
> Thanks,
> -Andy Torelli
>
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://www.phenix-online.org/mailman/listinfo/phenixbb
>   

Previous message: [phenixbb] Custom ligand - confirming the correct scattering	definition
Next message: [phenixbb] Confused by the output from Phenix autobuild jobs
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the phenixbb mailing list
George N. Phillips, Jr., Ph.D.
Professor of Biochemistry and of
     Computer Sciences
University of Wisconsin-Madison
433 Babcock Dr. Madison, Wi 53706
Phone/FAX (608) 263-6142



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20100215/9ffde221/attachment-0002.htm>