Skip to content

Latest commit

 

History

History
142 lines (118 loc) · 5.61 KB

ChangeLog.md

File metadata and controls

142 lines (118 loc) · 5.61 KB
Raw

Changelog of DockStream

1.0.0 - 2021-08-02 (RC)

Added

  • Minimal environment definitions.

Fixed

  • Fixed bug in Schrodinger/LigPrep sublist generation ('0' was interpreted as '1').

Internal

  • Added additional unit test for Schrodinger/LigPrep CLI application.

0.2.2 - 2020-04-30

Added

  • Extended input overwrite capabilities for docker.py (CSV specification).
  • New parsing capability to load SDF input with DockStream nomenclature to ensure enumerations are properly assigned.
  • Exposed additional command-line parameters for Schrodinger/LigPrep.
  • Added execution of Schrodinger/LigPrep` on AWS ("relative paths" bug on Schrodinger's side - workaround).

Fixed

  • Hotfix to improve Corina result parsing.

Internal

  • Show-case of how to execute Schrodinger/LigPrep on AWS including a token guard.
  • Show-case of how to execute Schrodinger/Glide on AWS including a token guard.

0.2.1 - 2021-04-16

Added

  • Additional measures to protect agains failing ligands at the embedding stage.
  • Added OpenEye/OMEGA embedder.

Fixed

  • Hotfix to increase stability of RDkit ligand preparator.

Internal

  • Added unit test to cover changes in parameters for LigPrep when using EPIK.
  • Improved logging for embedding stage.
  • Reverted file system handling for Glide.

0.2.0 - 2021-03-31

Added

  • New backend Hybrid (OpenEye) added (will replace the API version shortly).
  • Added stereo-isomer support using RDkit.
  • Added stereo-isomer support using Corina.
  • Exposed Corina's "-d" options.
  • Added support for Ligprep's filtering capabilities.
  • Enhanced stability for parallel execution mode for Hybrid, Gold and Glide.
  • Benchmarking script added.
  • Analysis script added.

Fixed

  • Fixed instability with protonations when using RDkit ligand embedding.
  • Fixed logging bug within Ligprep/Schrodinger.
  • Hot-fix for problem with CSV write-out when no pose was accepted.
  • Improved logging for Ligprep/Schrodinger.

Internal

  • Introduced stereo-enumeration factory class.
  • Adde pydantic-based parsing.
  • Optimized log messages.
  • Removed most versions from environment specifications.

0.1.4 - 2020-11-03

Added

  • Added support for parallelization for Ligprep/Schrodinger embedding.
  • Added best_per_ligand write-out mode for conformers.
  • Added CSV-option for entry point sdf2smiles.py.

Fixed

  • Fixed bug for OpenEye backend, when max_compounds_per_subjob was set to 0.
  • Fixed bug in best_per_enumeration write-out mode for CSV results.
  • Fixed irregularity with Gold conformer ordering (fitness versus score).
  • Fixed naming bug in OpenEye backend.

Internal

  • Extended AutoDock Vina unit tests for tautomer compounds.
  • Small update of Ligprep/Schrodinger unit tests.
  • Updated docking backend unit tests to fully cover enumerations.
  • Added unit test for Glide/Schrodinger constraints.

0.1.3 - 2020-09-29

Added

  • Added OpenBabel/AutoDock Vina target preparation (generating PDBQT files).
  • Added OpenBabel/AutoDock Vina box extraction (based on XYZ coordinate ranges of template ligand).
  • Added AutoDock Vina backend and result parser.
  • Added possibility to specify binary path for external binary executions.

Fixed

  • Fixed issue with logfile logging for Glide when time was exceeded.
  • Fixed issue with OpenBabel binary call when environment was not loaded.

Internal

  • Implemented hard over-write of POSE_OUTTYPE to be the only supported type.
  • Redesign of internal dictionary usage.
  • Addition of unit tests.
  • Added logging the version number (and started tagging versions).
  • Changed internal structure to be "package ready".
  • Improved the tag adding (and extended it for the ligand preparation step).

0.1.2 - 2020-09-09

Added

  • Integration of a "progress bar" to docking jobs.
  • Added support for parameter max_compounds_per_subjob to reign in (sub-)lists too long (especially with Glide).
  • Added option to only output the best scores per enumeration.
  • Added ligprep to available ligand embedding techniques.

Fixed

  • Fixed issues with "internal alignment" and integrated a fail-safe version.
  • Made all receptor paths lists rather than simple strings to streamline the interface prior to the implementation of ensemble docking.

Internal

  • Clean-up of "OpenEye" result parser.
  • Refactored result parsers.
  • Improved Gold feedback for docking.
  • Restructuring of internal "Ligand" handling.
  • Refactored "docker.py" entry point.
  • Added result parser output checks to respective unit tests.
  • Refactored some methods of the ligand preparation tools.
  • Updated example configuration files.

0.1.1 - 2020-08-17

Added

  • Added possibility to change the logfile path.
  • Added parameter to change the time limit per compound for Glide docking.
  • Added support to set a prefix for the output files.
  • Refactored and extended tagging system for all backends (adds "smiles" and "original_smiles" now).
  • Added option to only output the best poses per enumeration.
  • Added transformation support (using SMIRKS and OpenEye) for ligand preparation.
  • Added Glide/Schrodinger token guard.
  • Added support of "SDF" files as input.
  • Added support of arbitrary names to the parsing of CSV files as input.
  • Added "-debug" parameter to entry points.

Fixed

  • Critical fix for score aggregation if enumerations were used.
  • Made "Corina" embedding much more stable.
  • Several minor bug fixes in the logging write-out.

Internal

  • Added "ChangeLog.md".
  • Changed entry point structure (refactored and harmonized).
  • Fixed issue with "min"/"max" docking scoring directions (Gold/CCDC backend).