I am trying to get the difference electron density maps for residues with multiple conformations after I set the occupancy of their atoms to 0.

My main goal is to visualize the green positive density peaks that should appear for the atoms whose occupancy I set to 0.

However, when I run the phenix.refine command (doing all of this via cli, not gui), the output pdb file 

I will try to illustrate my process and problem with an example that works fine and one that doesn't, side by side, step by step.

1. I generate a pdb file that has a chosen residue set to occupancy 0. See 'pdb_1agy_res_SER_resID_120_del.pdb' and 'pdb_1agy_res_SER_resID_129_del.pdb'. As you can see, the former has SER 120 set to 0 occupancy and has one conformation, and the latter has SER 129 set to 0 occupancy for BOTH of its conformations. Both files are correct up to this point.
2. I then execute the following command:

phenix.refine pdb_del_file mtz_file main.number_of_macro_cycles=5 output.write_eff_file=False output.write_geo_file=False output.write_def_file=False refinement.input.xray_data.r_free_flags.generate=True --overwrite

• with mtz_file set to '1agy.mtz' (from the PDB) and pdb_del_file set to either 'pdb_1agy_res_SER_resID_120_del.pdb' and 'pdb_1agy_res_SER_resID_129_del.pdb'.

• This results in 2 files pdb_1agy_res_SER_resID_XXX_del_refine_001.mtz and pdb_1agy_res_SER_resID_XXX_del_refine_001.pdb. Comparing these pdb files, you can see that the occupancy for SER 120 remains set to 0 (which is what I want), but the occupancies for the 2 conformers for SER 129 have been filled in by phenix (not what I want) somewhere during the refinement process. As a result, I get the desired difference peaks for SER 120, but not for SER 129 (which seems to be averaged out).

Other than deleting the 2nd conformation for SER 129, is there a way to solve this problem, preferably via command line arguments/flags?

Thanks,
Alex