I don’t believe there are any tutorials/documentation which covers it directly, but it should be possible.
The key issue is to generate the params files for the ligand you want to use during the docking. If you look at the protein-small molecule docking tutorials, they should cover this. (The ligand preparation process is the same, regardless of the downstream protocol.) The one “trick” here is that for protein-protein docking, you will also need centroid parameters. The molfile_to_params.py script can generate these for you, or alternatively recent versions of Rosetta should be able to automatically generate the centroid parameters from the fullatom ones for most ligands.
Once that’s done, simply load the ligand parameters into Rosetta (same way as done with small molecule docking), and use the protein-protein docking code more-or-less as normal. The one trick is that you have to be careful about your chain/jump/docking partner specifications, so you’re docking the correct complexes. (But this isn’t any different if you’re docking a protein dimer to another protein dimer.)
.
