OEMedChem TK 1.0.4

New features

• A major overhaul of the fragmentation internals of OEMCSFragDatabase now allows an arbitrary number of fragmentation cuts. Additionally, an experimental capability for performing limited ring bond fragmentation cuts has been provided. Because of the combinatorial explosion of fragments generated with increased fragmentation cuts, care should be used when determining fragmentation cut limits. The enhanced ability to do arbitrary N-cut fragmentation makes the MCS results from the fragmentation approach much closer to an OEMCSSearch mapping result, with better overall performance than a mapping approach. To specify the fragmentation cut limits, see OEMCSFragDatabaseOptions::SetFragLimits, OEMCSFragDatabaseOptions::GetMinFragLimit, OEMCSFragDatabaseOptions::GetMaxFragLimit, OEMCSFragDatabaseOptions::SetRingFragLimit, and OEMCSFragDatabaseOptions::GetRingFragLimit.

• A new method, OEMCSFragDatabase::IsIndexed, has been added to determine whether a particular molecule index has been registered in the database. Depending on the indexing option settings, a particular molecule may be rejected from indexing. This method allows that condition to be interrogated.

• A new method, OEMatchedPairAnalyzerOptions::HasIndexableFragmentHeavyAtomRange, has been added to allow an options object to indicate that a min-max range has been specified via OEMatchedPairAnalyzerOptions::SetIndexableFragmentRange.

• The OEMatchedPairAnalyzerOptions options class now supports the operator== comparison to more easily determine the equality of two options objects.

• OEConfigureMatchedPairIndexOptions and OEConfigureMCSFragDatabaseOptions now support an -allfrags option to clear the min-max indexing constraint and to allow the indexing of all fragments. Indexing all fragments has indexing size and indexing performance implications for OEMCSFragDatabase and OEMatchedPairAnalyzer objects.

• The OEConfigureMCSFragDatabaseOptions function now supports the newly added -minfragcuts, -maxfragcuts, and -ringcuts options.

• OEReadMatchedPairAnalyzer now supports the ability to merge indices generated from segments of a large set of input structures. This capability can be used in a parallel processing approach to increase the throughput of the time-consuming structure indexing phase.

Preliminary Fragmentation API

• New Preliminary API methods have been added to allow the interrogation of intermediate fragmentation output and internal index information from the OEMCSFragDatabase object. See OEMCSFragDatabase::MoleculeToCores, OEMCSFragDatabase::CoreToMolecules, OEMCSFragDatabase::CoreToMoleculeCount, OEMCSFragDatabase::GetMaxCoreMolecule, and OEMCSFragDatabase::GetMaxCoreMoleculeCount.

• A new example has been provided to demonstrate using the new experimental fragment API of OEMCSFragDatabase. See MCSFragOccurrence.py.

Minor bug fixes

• A bug that resulted in the return of [R1][R2] fragments in the Matched Molecular Pair transforms returned by OEMatchedPairGetTransforms has been fixed.

• The OEMCSFragDatabase::GetScores has been changed to make the API more consistent with OEFPDatabase. The end range argument is now exclusive instead of inclusive when a non-zero value is specified.

• The OEMCSFragDatabase::GetScores and OEMCSFragDatabase::GetSortedScores methods previously returned an arbitrary OEMCSMolSimScore when different identified cores had the same score (in particular for OEMCSScoreType::AtomCount scoring). This has been fixed.

• The undocumented internal debugging method OEMedChem::OEMCSFragDatabase::Dump has been removed from the public API.

Python-specific changes

• Minor changes to some examples were made to synchronize the behavior and output with other wrapped language examples.

Java-specific changes

• Minor changes to some examples were made to synchronize the behavior and output with other wrapped language examples.

C#-specific changes

• Minor changes to some examples were made to synchronize the behavior and output with other wrapped language examples.

Documentation changes

• New methods have been documented in their respective classes.