The clean_pdb.py script was really written with the thought it would be used for monomeric proteins. As such, it will likely strip out non-protein residues like RNA.
Rosetta these days is actually pretty good at treating non-protein residues, so I might recommend just skipping the clean_pdb.py script. — This may, however, require you to do a bit of manual prep of the protein/RNA complex, manually deleting things like crystallographic additives or residues/regions of the protein you’re not interested in. But doing it manually should give you more control of the outcome.
If that’s not the issue (if your cleaned PDB contains all the residues you’re interested), it would help to get more information about the particular errors you’re seeing. (Copying-pasting the error message from the tracer output is a often useful.)
.
