I am trying to relax the pentameric 1SAC PDB file with its 10 calcium cofactors. I first removed other ligands from the file (ACY, acetic acid molecules) that I am not interested in, and left the calcium heteroatoms. I scored the PDB file using score_jd2.linuxgccrelease with the command lines flags -in:auto_setup_metals, -ignore_unrecognized_res, and -out:pdb. This worked well, and recognized the connectivity of the metal atoms with the appropriate residues. It also output a scorefile with a new scoring term called "metalbinding_contstraint" (scored using Talaris2014).
Then I tried went to relax the resulting PDB. I used the following command line flags:
.../relax.linuxgccrelease -s 1SAC_mBIO.pdb -relax:constrain_relax_to_start_coords -relax:coord_constrain_sidechains -no_optH false -optH_MCA true -flip_HNQ -ex1 -ex2 -use_input-sc -nstruct 10 -in:auto_setup_metals
In the tracer outpout, I can see that the metal/residue connections are again being recognized and that the "metalbinding_constraint weight was set to 1.0" for the Talaris2014 scoring. The relax then exits due to the following error:
AngleConstraint::score: warning: 0-length bonds!
[ERROR] EXCN_utility_exit has been thrown from: src/core/scoring/constraints/AngleConstaint.cc line: 231
I read on the "Work with Metalloproteins" page (https://www.rosettacommons.org/docs/latest/rosetta_basics/non_protein_residues/Metals) that the flag -in:auto_setup_metals will auto create angle constraints for metal ions, and I should not have to specify the metal using the -extra_res flag and CA.param file since most metals are already in the fa_standard residue set.
It seems that I missing something. Any help would be appreciated!