[phenixbb] Phenix.refine question

I am refining a small protein crystal model with 1.3 A data, there is a Tyrosine that won't stay put where it looks to me it should go (see below in it's original position, in the negative density; I moved the ring into the positive Fo-Fc density). After I move it into the positive density, save the coordinates, and do another refinement with the new coordinates, I find it right back in the negative density, even if I only refine ADP or OCC. Is there some trivial reason this is happening, or can I tell it to not refine that residue? https://docs.google.com/document/d/1RkZSH312q2Pimxb9gKLjibm1XbpsVpyzmrX4wGxr... THanks Laurie Betts