> Fit only works with objects that are identical in terms of atom number, > chainID, atom name and what not. This is not the case for a protein on > one side and a protein plus ligand on the other.
I've gotten away with some differences between chains. Sometimes you need to force it to fit using just a subset of residues. FYI, 'align' will actually result in a different superposition than 'fit' if you use the default settings, because it refines the alignment and throws away outlying atoms. Calling it with refinement off: PyMOL> align protA, protB, 0, 0 will result in an alignment identical to the one produced by 'fit', as far as I can tell. (I find for my purposes the align command with refinement on is better, but the RMSD will be different and not global.)
