Hi Kyle,
I have been performing standard refinement of atomic coordinates and B factors (ADP) in phenix.refine using NCS and bulk solvent correction without anisotropic scaling. The program has been returning a handful of atoms with B factors of zero after refinement, and I'm not sure why. I tried using adp.set_b_iso to reset all B factors before refinement but some still come back down to zero. Has anyone else noticed something similar, or has a suggestion on what to do?
is it a water oxygens? Typically, if a water oxygen gets placed into a heavier atom density (ion, metal) then its B-factor refines to much smaller value, sometime close to zero. So I would investigate this possibility. If it's not the water, then I don't know what's happening. And to tell more I would need to look at the data and model and probably run some refinement myself. Pavel.