[phenixbb] SA omit map

Nathaniel Echols nechols at lbl.gov
Wed Jul 6 09:53:45 PDT 2011


On Wed, Jul 6, 2011 at 6:30 AM, Tjaard Pijning <t.pijning at rug.nl> wrote:
> The resulting map (/OMIT/resolve_composite_map.mtz) shows density for both
> the protein and the ligand. When I feed this into the CCP4 module FFT to
> generate an nF1-mF2 map (with n=1 and m=1) I still get density for both the
> protein and ligand. However I need only the ligand Fo-Fc density.
> What is the right way to do this ?

Pavel is correct, it is easiest to do this with phenix.refine instead
of AutoBuild.  An alternative to the commands he suggested is simply
to delete the ligand from the input model and then run refinement as
usual (with simulated annealing).  This gives you total control over
what is omitted; I think AutoBuild may carve out a box around the omit
atoms.  The Fo-Fc map coefficients will still cover the entire unit
cell, but if you have correctly placed all of the protein atoms and
obvious solvent, it will mainly show your ligand.  Otherwise, you need
to use a different tool to limit how much of the map is displayed -
Phenix won't do this for you, at least not if you want to use
simulated annealing.  (We could add a separate tool for this, of
course - it's very easy to implement.)

-Nat


More information about the phenixbb mailing list