- Added
timeoutargument tocompare_moleculesin the PARITY method to control the MCS calculation time (default 40 seconds) - Added
validate_ligand_cif_categoriesandvalidate_mm_cif_categoriesto validate ligand and macromolecule CIF files before reading - Errors raised during sanitisation are now reported in the
errorsof the reader results - CIP labels are assigned to molecules after sanitisation (#36)
- Stereochemistry is assigned from the Ideal conformer when it is valid, falling back to the Model conformer only when the Ideal one is degenerate
- Physicochemical properties are calculated on the molecule without hydrogens, so bonds to terminal hydrogens are no longer counted as rotatable bonds
- RDKit logging no longer redirects the global standard error output
- Added
errorsfield toSanitisationResult, a list of"<error type>: <message>"strings sanitizeno longer converts bonds between metals and other atoms to dative bonds to fix valence issues- Ligand CIF readers raise
CCDUtilsErrorwhen_chem_comp.or_chem_comp_atom.categories are missing, or when a macromolecule mmCIF file is provided - Bound molecule reader raises
CCDUtilsErrorwhen_entry.,_atom_site.or_chem_comp_bond.categories are missing - Missing optional categories are reported as a single warning (
missing: ...) instead of one warning per category compare_moleculesraisesTimeoutErrorwhen the MCS calculation times out- Physicochemical property values (e.g.
NumRotatableBonds) differ from previous versions - Removed
preprocessable_categoriesfromccd_reader,clc_readerandprd_reader
- Support for RDKit 2025 and NumPy 2.0
- Uses both coordgen library and the default RDKit functionality to generate 2D depictions
- Enable parsing of a subset of CCDs from the Chemical Component Dictionary
- Added CCDC to UniChem resources
- Updated for RDKit 2023.09.6
- Removed the formal charge adjustment for valence issues, instead uses Dative bonds
- Uses Zero bond type (dotted lines) for depiction of coordinate bonds
- Changes Dative bonds to single bonds for inchi calculation
- Corrected the format of PDB files
- Corrected the header of SDF file
- Removed smile from SubstructureMapping and added mol
- Added clc_reader and clc_writer to read and write Covalently Linked Components (CLCs)
- Added prd_reader and prd_writer to read and write Protein Reference Dictionarys (PRDs)
- Added boundmolecule_cli, a pipeline to infer CLCs from PDB model files
- Replaced pdbecif with gemmi for mmcif parsing
- Support for rdkit 2022.09.x
- Component representation of Covalently Linked Components (complex multi-component ligands typically represented as individual components represented by individual CCDs)
- Removed ccd_cif_dict (dict) property of Component. Please use ccd_cif_block (gemmi.cif.Block) to access data from mmcif file
- A lot of minor bug fixes and code improvements
- New templates for 2D layouts
- Support for rdkit 2021.03.x
- Add support for UniChem mapping.
- Add bond information to the SVG decomposition.
- Allow PARITY method to be atom/bond specific.
- Improve and extend physchem properties.
- Enhanced CIF export (physchem, scaffolds, fragments, 2D, mapping).
- Few improvements to match newest RDKit version (breaking changes).
- Add SVG decomposition in the SVG format.
- Protein-ligand interaction pipeline moved to separate repository.
- Add basic properties calculation (Abhik).
- Extension and improvements of the Fragment library.
- Introduce CoordGen from RDKit.
- Refactored pdbeccdutils.core (breaking changes).
- Scaffolding (Abhik).
- Add protein-ligand interaction pipeline.
- PDBeChem pipeline.
- Support for PARITY method.
- Introduce EKTGv2() method for 3D conformer generation.