OEMedChem TK 0.9.5¶

New features¶

The matched pair analyzer now supports the identification of matched molecular pairs that differ by heavy-atom and hydrogen-atom substitutions. This additional indexing activity returns significantly more matched molecular pairs than the previous version. The hydrogen-atom indexing activity has been enabled by default. To perform heavy-atom indexing only, remove the OEMatchedPairOptions_IndexHSites option from OEMatchedPairOptions_Default and set the desired option(s) explicitly via OEMatchedPairAnalyzerOptions.SetOptions.
The matched pair analyzer now supports single bond insertion differences between matched molecular pairs. Previously, a minimum change of 1-atom was identified in the substituent change between the matched pairs.

OEMatchedPairOptions_IndexHSites is a new option that enables hydrogen-atom indexing support to the default indexing activities.
OEMatchedPairTransformExtractMode_SuppressNodeSort is a new option that suppresses the internal sorting of index nodes prior to extraction of transformations and matched pairs for debugging and analysis of indexing problems.
OEMatchedPairTransformExtractMode_AddMCSCorrespondence is a new extraction option that returns the maximum common substructure correspondence between the matched molecular pairs in the form of atom-mapping information applied to the MCS cores.
OEMatchedPairTransformExtractMode_AddAttachmentPts is a new extraction option that requests the addition of Rgroup attachment points on the returned matched molecular pair transforms.

A number of bugs related to the identification of the maximum common substructure for matched molecular pairs have been corrected.
The ChEMBL solubility data has been re-indexed to capture indexing fixes, potentially impacting the results returned from OEApplyChEMBL18SolubilityTransforms.
An internal index node sort now includes the number of fragmentation cuts in the sort criterion to preferentially return single > double > triple cuts for the largest maximum common substructures.
A significant refactoring of internal activities has been performed to prepare for additional public API exposure in future versions.