We'll likely have some need for structural alignment and other interesting protein features so we should a PDBx parser. See poly issue [#297](https://github.com/TimothyStiles/poly/issues/297)