It is often easier to specify CONECT records for a custom amino acid than it is to get all the atom names right. The information needed to match on CONECT records is present in a ResidueDefinition. So it should be simple to match residue definitions against connectivity when atom names do not produce a match (before filtering on linkages). This would seamlessly improve the user experience of trying to prepare a custom residue, as when they have CONECT records they would not need to synchronize the atom names between the PDB and residue database, without changing the API at all. Users could choose whether to prepare the PDB file to match on atom names or connectivity.