Closed brindakv closed 2 years ago
python-ihm tries to be pretty tolerant with its input. Are you still using util/make-mmcif.py
in your pipeline? If so, easy to add a check there for duplicate atoms.
Right. But the dumper also writes out coordinates for multiple models with the same model number.
Sure, we could check in the dumper instead, I guess that makes sense. We already check for other wonkiness (e.g. trying to emit atoms for a seq_id that isn't in the representation) so that would be a natural place to do it.
We recently received a deposition where the uploaded mmCIF file had ten models in the
atom_site
table. However, the_atom_site.pdbx_PDB_model_num
column had the same model number for all ten models i.e., there were multiple sets of coordinates for the same atom in the same model.python-ihm
did not throw out an error while reading this mmCIF file.