FastROCS TK 1.8.3

New features

• Alternative starting points in FastROCS TK have been extended to include two additional API points: OEFastROCSOrientation::Random and OEFastROCSOrientation::Subrocs.

  • OEFastROCSOrientation::Random now randomly orients the database molecule N times. By default, N is set to 10 with a maximum allowed translation of 2 angstroms. Random alternative starts can be customized using the OEShapeDatabaseOptions::SetMaxRandomTranslation, OEShapeDatabaseOptions::SetNumRandomStarts, and OEShapeDatabaseOptions::SetRandomSeed methods. The default random seed is time-dependent.

  • OEFastROCSOrientation::Subrocs now performs optimizations at every heavy atom of the molecule with the most heavy atoms, regardless of which molecule is the query and which is the database molecule. At each translation, 4 inertial poses are optimized, resulting in a total of [4 * number of heavy atoms of the heaviest molecule] optimizations.

  See OEFastROCSOrientation and OEShapeDatabaseOptions::SetInitialOrientation for more information.

• The following example scripts have been reworked to enable access to additional features of FastROCS TK:

  • ShapeDatabaseServer.py

  • ShapeDatabaseClient.py

  • ShapeDatabaseClientHistogram.py

  • ShapeDatabaseProxy.py

  Tversky sampling, shape-only searches, and basic alternative starting points can be accessed via the client without having to rerun or modify the server/proxy scripts. Users can submit a query as usual from the client adding the relevant flags. The new flags are:

  • --tversky

  • --shapeOnly

  • --alternativeStarts [inertialAtHeavyAtoms|inertialAtColorAtoms|subrocs|random [N]]

  Warning

  These scripts are not backwards compatible. The behavior of running server/proxy/client scripts with mismatched versions is undefined. In order to use the new client flags, ensure that the server and proxy scripts are also updated to this version.

• Warnings that are thrown for Tversky and Tanimoto scores greater than 1.0 have been simplified. Messages showing the actual Tversky/Tanimoto calculation are now Debug-level messages.

Minor bug fixes

• OEDBTracer::GetTotal now correctly accumulates the total when using the same OEDBTracer object for multiple queries.

• OEShapeDatabaseOptions::SetLimit now correctly handles the scenario where the limit is greater than the maximum scores available. If the limit is greater than the possible number of scores able to be returned, a warning will be thrown and the limit will fall back to: number of conformers * OEShapeDatabaseOptions::GetMaxOverlays.

Python-specific changes

• The example script karma.py in FastROCS TK examples has been renamed to the more descriptive SphereExclusionClustering.py.

Documentation changes

• A new tutorial, Tutorial 2, has been created for preparing a database to take advantage of faster OEShapeDatabase::Open times.