Rosetta is somewhat capable of considering your ions in fixbb. To get the ions working, you need parameters files for them. Look in the folder rosetta_database/chemical/residue_type_sets/fa_standard/residue_types/metal_ions to see which metals Rosetta already has programmed. You may need to write your own parameters files if none of these are the right elements. Also check in rosetta_database/chemical/residue_type_sets/fa_standard/residue_types.txt to make sure your ions are turned on (not # in front of the line).
Secondarily, you need to make sure the ions in your input match the spelling and spacing the parameters file expects. For demonstration, I’ll substitute spaces in the PDB for underscores: if you have “MG_”, and it doesn’t work, try “_MG”. You may need to swap HETATM for ATOM.
Finally, the quality of Rosetta’s consideration is somewhat suspect. Rosetta doesn’t know about the orbitals and bonding patterns of ions, it will consider their van der Waals radius and that’s about it. You aren’t using solvation and the standard scorefunction’s pseudo-electrostatics (fa_pair) ignores ions. If you want beautifully tetrahedrally-coordinated zinc, you aren’t going to get it without some heavy-duty additions to the scorefunction. If you just want atoms taking up space to prevent some ion-dependent conformation from collapsing, it oughta work.
.
