[phenixbb] racemic ligand refinement

All, I need to refine a ligand that consists of two stereoisomers in my active site. I have decent density and have created models for the r and s forms of the guest with separate cif files. They have been fit independently into the active site density. My questions are; How to show these in the pdb file. Separate chains? Separate amino acids in a given chain? How do I keep them from interacting with one another during refinement? Creating an alt-position seems obvious but I'm not certain how to apply the different stereo restraints in that case. I'm sure someone has solved this before but I can't seem to find that information. Thanks, Nic Steussy Purdue University.