[phenixbb] picking Rfree set

[Ralf W. Grosse-Kunstleve]

> 2008/1/22, David Garboczi <dgarboczi at niaid.nih.gov>:
> > So, for instance, when addition of a ligand to a P212121 cell (1 mol)
> > changes the symmetry to P21 (almost same cell, but 2 mol), the
> > selection of the Rfree set from the P21 data by phenix.refine will
> > take the higher almost symmetry into account?
> >
> > Dave
>
> yes, it will.
>
> This is useful for when the data is twinned.
>
> P

But watch out for large deviations from the orthorhombic symmetry. The
tolerance is controlled by the r_free_flags.lattice_symmetry_max_delta
parameter. The default is pretty generous (5 degrees) but you should
check that that's enough for your case.

You can use the
  iotbx.lattice_symmetry
command with --delta=5 to verify that you get the orthorhombic
symmetry starting from your monoclinic cell.

Ralf