I have a strange problem regarding SymDock:
I did symmetry Rosetta previously using RosettaScrips 3.4. And it worked fine so far. Since I want to do docking on that symmetric molecule and - at least to my understanding - the Docking is not ported to Scripts for symmetry mode, I choose to use SymDock.
When I try to run it with exactly the same .pdb .sym and .params as in Scrips (which worked fine!) I get "
protocols.simple_moves_symmetry.SymDockingInitialPerturbation: initialize_rigid_body_dofs: true
ERROR: unrecognized atom_type_name COO
ERROR:: Exit from: src/core/chemical/AtomTypeSet.hh line: 103
"COO" is present in my ligands .params file, but should also be present in a lot of other residues ... so, I do not get the meaning of the error, since regarding the extra-res readin everything worked perfect in RosettaScripts.
I attached all relevant files and the full output.
Thanks a lot ... I suppose its stupid fault of mine ;-)
I've got this error before. It's because atom type COO is not included in 'centroid' stage of the simulation. You can find the allowed atom types in rosetta_database/chemical/atom_type_sets.
I worked around this issue by save a separate ligand params for centroid stage by changing 'COO' atoms to 'CObb' atoms (CObb is allowed in centroid stage, but it has larger volume and different orbitals compared with COO.) Then in the flags file, I used -extra_res_fa and -extra_res_cen to read in params file for full atom and centroid stages.
I'd also like to know if there's a better way to solve this.
When you create a params file with molfile_to_params, you can give it the command line flag "--centroid" to have it create a centroid-mode params file in addition to the full atom mode params file.
"--centroid" with molfile_to_params.py works like a charm.