Need association between comparative model asym_ids and sphere model asym_ids about ihmcif

[tomgoddard]

The mediator.cif cr_mid_fullmed10.pdb has chains D,G,I,J,U,Z and is referenced as dataset_list_id=2 by five starting models which have asym_id H,I,J,K,L.

[tomgoddard]

Is it assumed that the sequence numbers for dataset_list models referred to in the starting_model_details table exactly match the sequence numbers used by the sphere models (given by seq_id_begin and seq_id_end in the starting_model_details table?

If the sequence numbers are not required to match, then a correspondence between experimental/comparative input models and the sphere model sequence numbers will be needed.

[brindakv]

In the updated dictionary, _ihm_starting_model_details.starting_model_auth_asym_id and _ihm_starting_model_details.starting_model_sequence_offset address this issue.

[tomgoddard]

I think we need the starting_model_db_auth_asym_id field in addition to the new starting_model_auth_asym_id field in _ihm_starting_model_details. The db_auth_asym_id tells which change of a database entry was used in a comparative alignment. The auth_asym_id tells which chain in the resulting comparative model file corresponds to the asym_id in the sphere model. In the case of nup84 the comparative model asym_ids match the sphere model asym_ids, but in the case of mediator.cif comparative model (cr_mid_fullmed10.pdb the asym_ids don't match the sphere model asym_ids. We were able to repurpose db_auth_asym_id to match these only because that comparative model did not list database templates, so the db_auth_asym_id was not being used.