|
|
|
# Development documentation: use RDKit to write sdf file for PDB CCD
|
|
|
|
*28 May 2017*
|
|
|
|
* previous work [RDKit molecule from PDB CCD definition](rdkit-from-ccd) now extend to write sdf files.
|
|
|
|
* start with a stub implementation and unit tests [Commit 5773ff0e](https://gitlab.com/pdbe/ccd_utils/commit/5773ff0e2b3f560ad344777452a567a88e57cf45)
|
|
|
|
* initial implementation [Commit 7eb159fe](https://gitlab.com/pdbe/ccd_utils/commit/7eb159fedf4cc0cea5b3c53b92ca24995433664b)
|
|
|
|
* writes sdf files with the "ideal" coordinates and hydrogen successfully
|
|
|
|
* these sdf files can be read by Mogul and the atoms are labelled with their PDB atom name. For instance reading `tests_out/ATP.ideal_withH.sdf` and selecting atom labels on results in:
|
|
|
|
* 
|
|
|
|
* note that atoms are labelled according to the atom order in sdf that reflects the order in the input ccd cif. I thought that Mogul read the alias information but not in this case. Test was running an old Mogul version need to check with the latest. But actually it does not really matter.
|
|
|
|
* pymol can also load although it does not get atom order or alias
|
|
|
|
* obabel is confused by the alias but never mind.
|
|
|
|
* many things left to to do
|
|
|
|
* is this a sdf or a mol file?
|
|
|
|
* set sensible title for the molecule in the file
|
|
|
|
* what about the method line
|
|
|
|
* need option not to write hydrogen atoms
|
|
|
|
* need option to write exptl coords rather than ideal |
|
|
\ No newline at end of file |
|
|
|
moved to issue #3 |
|
|
\ No newline at end of file |
