Hi Simon, providing selection alone for occupancy refinement is not enough to actually tell program to refine occupancies. For that you really need to have "occupancies" in "strategy". If you don't want to refine occupancies of some atoms that would be refined by default, then yes, you need to use "remove_selection" syntax, just like you found out in the docs. Pavel On 5/15/12 4:49 AM, Simon Kolstoe wrote:
Thanks Pavel,
So if I turn on the overall occupancy strategy and then set the occupancies of the atoms I want to refine at 0.9 it works. HOWEVER it also refines the occupancies of atoms on a different chain that I want to fix at 0.5 (I have a ligand that crosses a symmetry axis). So what command do I use to turn off the occupancy refinement for the other chain? Is it something like:
occupancies.remove_selection = chain Y
(from the phenix.refine documentation page).
Thanks,
Simon
On 15 May 2012, at 10:56, Pavel Afonine wrote:
Hi Simon,
one of possible ways: just change the starting occupancy values to something between 0 and 1, for example, set them to 0.9. That will be a sign for Phenix to refine their occupancies.
Yours way should work too (though I don't know all the details, including what's in .def file), but may be you try resetting occupancies first?
Pavel
On 5/15/12 2:27 AM, Simon Kolstoe wrote:
Hi phenixbb,
I'm wanting to refine the occupancy for just a few Cadmium atoms in a structure but not refine any other occupancies. I've used the following command however the occupancies haven't refined:
phenix.refine refinement_006.def occupancies.individual="chain X and element CD2"
and the cadmiums are in the pdb file as:
ATOM 8740 CD+2 CD2 X 1 -81.043 91.612 65.218 1.00 11.63 X CD2+ ATOM 8741 CD+2 CD2 X 2 -78.111 92.072 62.948 1.00 11.15 X CD2+ ATOM 8742 CD+2 CD2 X 3 -85.283 28.170 80.931 1.00 12.01 X CD2+
Thanks,
Simon
phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb