I am trying to run Local Relax on an ADP-bound protein structure. I cleared the first hurdle to make a ligand params file, and the program runs fine and includes the ligand. I made the params file by taking ADP from a related structure, converted it to mol2 using iBabel, and made the params file using molfile_to_params.py
However, during all of my runs the ADP ligand totally loses planarity in the nucleotide base, and I end up with unrealistic ligand conformations. I think the most likely cause is that Rosetta is somehow not recognizing aromatic bonds.
I've double checked the params file and all of the bonds in the nucleotide ring are correctly labeled as aromatic.
I'm hoping that I've made a very obvious mistake, so please point it out!