Currently, atomworks reads in obsoleted CCD codes.
This is particularly noticeable with the LIG code, which some pipelines output. (e.g. OpenFold3 with SMILES ligands and PDB output types the residue as LIG). When read in, atomworks does not interpret the LIG residue as an arbitrary ligand, but rather as the historic 3-PYRIDIN-4-YL-2,4-DIHYDRO-INDENO[1,2-.C.]PYRAZOLE, generating spurious atoms which aren't in the input.
Ideally, atomworks should by default not use the CCD definitions for components with the _chem_comp.pdbx_release_status of OBS, treating them as if they were a component/residue identifier not found in the CCD (e.g. just like it does with L:G)
At the very least, it should avoid loading the wwPDB CCD definitions of any of the codes which the wwPDB has reserved as placeholders, namely 01 - 99, DRG, INH & LIG (DRG, INH, LIG & 98 all have obsoleted entries the new placeholder usage will shadow.)
_(The other placeholder specifications of UNK, UNL and UNX should also be appropriately special-cased if they aren't already, though those aren't actually obsoleted and are present in the wwPDB. Though this likely means that their behavior has already been optimized.)
Currently, atomworks reads in obsoleted CCD codes.
This is particularly noticeable with the
LIGcode, which some pipelines output. (e.g. OpenFold3 with SMILES ligands and PDB output types the residue asLIG). When read in, atomworks does not interpret the LIG residue as an arbitrary ligand, but rather as the historic 3-PYRIDIN-4-YL-2,4-DIHYDRO-INDENO[1,2-.C.]PYRAZOLE, generating spurious atoms which aren't in the input.Ideally, atomworks should by default not use the CCD definitions for components with the
_chem_comp.pdbx_release_statusofOBS, treating them as if they were a component/residue identifier not found in the CCD (e.g. just like it does withL:G)At the very least, it should avoid loading the wwPDB CCD definitions of any of the codes which the wwPDB has reserved as placeholders, namely
01-99,DRG,INH&LIG(DRG, INH, LIG & 98 all have obsoleted entries the new placeholder usage will shadow.)_(The other placeholder specifications of UNK, UNL and UNX should also be appropriately special-cased if they aren't already, though those aren't actually obsoleted and are present in the wwPDB. Though this likely means that their behavior has already been optimized.)