Hi everybody!
Everytime when I (within coot) do "real space refine" two arginines that have two alternate conformations ..(interacting by hydrogen bonds); and then phenix.refine everything, both conformers move away. Therefore, when I open my mtz and pdb files from phenix.refine in coot, I'll have green blobs where there should be guanidium groups.

Can I tell phenix.refine not to move only those two residues?

It should be something like the following lines in the eff file?
sites {
      individual = None
      torsion_angles = None
      rigid_body = None
      remove_selection=resid 45
      remove_selection=resid 68
}

Thank you very much