[phenixbb] Refining anomalous groups

Nathaniel Echols nechols at lbl.gov
Wed Jan 9 12:38:25 PST 2013


On Wed, Jan 9, 2013 at 12:26 PM, Brennan Bonnet
<Brennan.Bonnet at lightsource.ca> wrote:
> I hope everyone had relaxing holidays to begin the new year.  I am back to work trying to refine anomalous groups using phenix.refine.
>
> I have managed to merge my files so that my exptl_fobs_phases_flags_1_with_hl_anom.mtz now contains F(+), SIGF(+), F(-), SIGF(-).
>
> Now I have a new problem which was not occurring before the break.  Usually I check the box for "Anomalous groups", then select "Modify selections for: Anomalous groups" which brings up a screen for atom selection and definition of f' and f''.  In the past, clicking on the View/pick button launches Coot for selection of the atoms.  Sadly, now clicking on the View/pick doesn't launch Coot and instead crashes Phenix.

What OS (including version) are you using, which Phenix installer did
you use (including the exact package name), and could you please send
me any error messages printed to the console?  The built-in graphics
is spectacularly difficult to get working on some Linux systems, for
reasons I still don't understand.  However, I don't think those
buttons ever launched Coot - maybe something that looks superficially
similar, perhaps, but it's obviously stopped working on your computer.

I do have a suggestion for working around this, but it's a pain to
use: a few weeks ago I added an option to set up the anomalous groups
automatically.  You have to choose the "Edit parameters and run"
option when starting phenix.refine, then look for this parameter
(heavily condensed):

refinement {
  group_anomalous {
    find_automatically = False
  }
}

Change that to True, and every atom heavier than P will be treated as
a separate anomalous group.  (This is not necessarily the optimal way
to do it, which is why it's hidden right now, but it worked well for
the tests I've done so far.)

-Nat


More information about the phenixbb mailing list