It would seem very useful to allow _entity_poly_seq.seq_id as a pointer to _atom_site.label_seq_id to allow the alignment of chain sequences in a less cluttered environment than the atom_site list, especially for cases where a published nomenclature has been used for residues for which coordinates have not been obtained. If added to the key for entity_poly_seq, multiple alternate conformations and nomenclatures could also be aligned, provided a pointer to _atom_sites_alt.id (say _entity_poly_seq.alt_id) and, perhaps an _entity_poly_seq.alt_nomenclature_id were added to the key. It would be very convenient to have all the alternate residue names gathered in one cleanly parse-able list. -- H. J. Bernstein