Multiple Metal Sites in Autosol
??Hello All, ?I have a question regarding the metal assignments when using Autosol to process SAD datasets. I am new to field and I want to make sure that I am doing things correctly. My enzyme should contain 2 metal sites, one with Cu and one with Fe. We collected data at the Fe-edge, the Cu-PreEdge, and the Cu-Edge for a comparison. When I tell Autosol to search for 1 Cu site, using the Cu-Edge data, and another heavy atom Fe, it finds 33 sites all Cu with 2 being Fe. Likewise, using the same dataset, when I include S as an additional anomalous scattering moiety, it will find 1 Cu in the correct position and 32 S atoms--my protein does not contain 32 S atoms. Finally, when I include every possibility that was in my crystallization buffer, this includes a fairly high concentration of MgCl2, and I tell it to search for 1 Cu site with the additional heavy atoms being Fe, S, Mg, and Cl, it seems that it is able to assign the Cu correctly, however it now gives Mg coordinated with Cu instead of H2O molecules, and this is not possible. Should I include Mg and Cl in the search? And S? Also, why would it not be finding the Fe? Furthermore, and most importantly, why is it assigning so many heavy atom sites? I am having similar problems with the other two data sets as well. Thank you very much in advance! Regards, Ryan Gumpper
Hi Ryan, Within AutoSol, Phaser is using the real and imaginary parts of the scattering factors for these various metals to guess what atom goes where. For each site, it compares the agreement with the data for each atom type and it picks the best one. This can help in phasing and in identifying what site is what metal, but it is not intended as a method for identification of metal ions. Note that the identification will depend on the resolution and quality of the anomalous data and the differences among the scattering factors for the various metals. So I wouldn't worry too much about what atoms are found. Phenix.refine (see http://journals.iucr.org/d/issues/2014/04/00/lv5059/lv5059.pdf) will do this identification for you much better after your structure is solved. I hope that helps! -Tom T On Jun 24, 2014, at 9:55 AM, Ryan Henry Gumpper wrote: Hello All, I have a question regarding the metal assignments when using Autosol to process SAD datasets. I am new to field and I want to make sure that I am doing things correctly. My enzyme should contain 2 metal sites, one with Cu and one with Fe. We collected data at the Fe-edge, the Cu-PreEdge, and the Cu-Edge for a comparison. When I tell Autosol to search for 1 Cu site, using the Cu-Edge data, and another heavy atom Fe, it finds 33 sites all Cu with 2 being Fe. Likewise, using the same dataset, when I include S as an additional anomalous scattering moiety, it will find 1 Cu in the correct position and 32 S atoms--my protein does not contain 32 S atoms. Finally, when I include every possibility that was in my crystallization buffer, this includes a fairly high concentration of MgCl2, and I tell it to search for 1 Cu site with the additional heavy atoms being Fe, S, Mg, and Cl, it seems that it is able to assign the Cu correctly, however it now gives Mg coordinated with Cu instead of H2O molecules, and this is not possible. Should I include Mg and Cl in the search? And S? Also, why would it not be finding the Fe? Furthermore, and most importantly, why is it assigning so many heavy atom sites? I am having similar problems with the other two data sets as well. Thank you very much in advance! Regards, Ryan Gumpper _______________________________________________ phenixbb mailing list [email protected]mailto:[email protected] http://phenix-online.org/mailman/listinfo/phenixbb
On Tue, Jun 24, 2014 at 9:15 AM, Terwilliger, Thomas C wrote: Note that the identification will depend on the resolution and quality of
the anomalous data and the differences among the scattering factors for the
various metals. So I wouldn't worry too much about what atoms are found.
Phenix.refine (see
http://journals.iucr.org/d/issues/2014/04/00/lv5059/lv5059.pdf) will do
this identification for you much better after your structure is solved. Mixtures of transition metals are a little tricky, and the same disclaimer
applies to phenix.refine. Cu and Fe may be different enough to permit
identification, but Cu and Zn are not even at atomic resolution. Ideally
we would incorporate multi-wavelength analysis in the ion placement but
this is a long-term project.
-Nat
participants (3)
-
Nathaniel Echols -
Ryan Henry Gumpper -
Terwilliger, Thomas C