• Docs »
  • Release Notes

Release Notes

Release Highlights 2018.Oct

Omega TK: Macrocycle

A new distance geometry-based, conformational space-sampling method has been added to Omega TK as a preliminary API. This approach opens up a new avenue for exploring conformer ensembles in Omega TK. Although the distance geometry-based sampling is slower than the traditional torsion driving-based approach, it works substantially better for ensemble generation of macrocyclic molecules. This sampling method can also be applied to linear and small ring-containing molecules.

../../_images/geldanamycin_macrocycle.png

Conformer ensemble subset for geldanamycin (PDB: 1YET) generated using the macrocycle method

FastROCS TK: C++ and Java Support

FastROCS TK previously had been available only in Python. This release extends support to C++ and Java. All classes, functions, and constants are wrapped and available for use. Several example scripts now have C++ and Java versions to showcase FastROCS TK functionalities. Additionally, users running FastROCS TK on Tesla GPUs will see a 10-30% speed-up over the previous version depending on the system.

Quacpac TK: Tautomers

This release adds several improvements to the tautomer functionality in Quacpac TK. Tautomer handling in molecular modeling and cheminformatics usually includes three important functions:

  • Generating a canonical protomer representation for duplicate-free storage
  • Generating a single tautomer suitable for visualization by chemists and modelers
  • Generating an ensemble of aqueous-phase tautomers suitable for molecular modeling

Example of generating an ensemble of biologically relevant tautomers

This release provides new API to cover each of these use-cases. The algorithm has also been improved to handle memory usage when a large number of possible tautomeric forms are present.

General Notices

  • The release cycle of OpenEye Toolkits has been changed. See the Release Cycle section for more information.
  • This is the last release to support GCC 4.4 on RHEL6. For future toolkit releases, GCC 4.8 will be the oldest targeted compiler.
  • This is the last release to support OS X 10.11.
  • FastROCS TK no longer supports Nvidia Tesla, GEForce, and Quadro cards with a Compute Capability older than 3.0. This includes all GPUs with Fermi architecture (e.g., Tesla C2050 and GEForce GTX 430). If you are running one of these older GPUs and cannot upgrade your hardware, do not update to this version of FastROCS TK.
  • Spruce TK now requires a specific license. For information on Spruce TK licensing, please contact sales@eyesopen.com.
  • Improvements have been made in the Python 3 toolkits with regard to wrapping optimization. This results in Python 3 performance being comparable to Python 2.

Detailed Release Notes

OEChem TK 2.1.5

New features

  • New OEHasExplicitHydrogenExpression and OEHasImplicitHydrogenExpression functions have been added.
  • The OEPrepareSearch function now takes a parameter that determines whether to adjust the hydrogen of a target molecule based on the presence or absence of implicit and explicit hydrogens on the query molecule of the substructure search.
  • Both OEGetSubtree and OEGetSmallestSubtree functions now take an optional max depth parameter that allows partial traversal of subtrees. For example, the parameter allows identification of the subtree of a proline residue without traversing back onto the main chain. Similarly, it is possible to get the subtree of a cysteine residue involved in a disulfide without traversing across the disulfide bond and back on the main chain.
  • Two new constants, OEIFlavor.MDL.SuppressImp2ExpENHSTE and OEOFlavor.MDL.SuppressImp2ExpENHSTE, have been added with related MDL input and output formats (e.g., SDF, RDF) to allow suppression of explicit MDL stereogroup generation for implied stereocollection membership. By default, marked stereocenters not explicitly included in stereocollections are added to their implied stereocollection membership based on the chiral flag setting.
  • A new OEGroupType.UserDefined type has been added to OEGroupBase to support serialization and deserialization via the OEB format.
  • A new OEIsBridgeHead functor has been added to identify bridge head atoms in a molecule.
  • A new OERingBondCount function has been added that returns the number of bonds connected to an atom that are part of a ring.

Major bug fixes

  • The FASTA reader has been fixed to support reading multiline FASTA strings.
  • Generic data attached to OEGroupBase is no longer lost when the molecule is copied.
  • Generic data attached to OEGroupBase is no longer lost when the molecule is added to another molecule by invoking the OEAddMols function.
  • Relative stereo (e.g., 1,4-disubstituted cyclohexanes) information is now preserved on input, ensuring that the conformers generated will respect the input stereochemistry. Previously, such relative stereocenters would be cleared during activities to clear other illegally marked stereochemistry.
  • The V3000 MDL format readers called by OEReadMolecule now support noncontiguous object indices. It previously had been assumed that such indices were contiguous, but recent third-party writers have demonstrated that this assumption is no longer valid.

Minor bug fixes

  • The OEGenerate2DCoordinates function now returns false when it fails to generate coordinates that are consistent with the atom/bond stereo of a molecule.
  • A memory corruption issue in the InChI generation code that previously required a check for hydrogen isotopes beyond [3H] will no longer generate InChI for these unrealistic hydrogen isotopes.
  • Malformed SD bond lines that lacked both atom endpoints had previously been ignored. OEReadMolecule now fails for these types of format errors.
  • OEUniMolecularRxn internal validations have been relaxed to allow reaction transformations that create a bond between two separate fragments.
  • OEGetHybridization has been repaired to return OEHybridization.sp3d and OEHybridization.sp3d2 for high valence halogens. Previously, this function had only returned OEHybridization.sp3 for these cases.
  • OEUncolorMol has been augmented to remove stereogroups when stripping atom stereochemistry. Additionally, a new option, OEUncolorStrategy.RemoveGroupStereo, has been added to support the explicit stripping of stereogroups.
  • OEAnd and OEOr have been augmented to support using OEGroupBase types to allow a more complex selection of MDL stereogroup types via custom predicates.
  • A new copy constructor and new assignment operator have been added to OERotMatrix.

Python-specific changes

  • Type comparisons between OEBase and non-OEBase objects have been fixed to ensure that an exception is longer thrown.
  • A new __str__() function has been added to OEResidue to provide a convenient string representation.

Java-specific changes

  • A new OELibExtractor.GetTmpDirName() method has been added to return the temporary directory used for the extraction of native libraries from the JarFile and the extraction lock file. This directory location can potentially be used to clean up extraction locks that have been left behind due to inadvertent or explicit Java process kill activities.

Documentation changes

OEBio TK 2.1.5

New features

Minor bug fixes

  • It is now possible to set an alternative rotamer on proline residues.

C++-specific changes

  • OEBio TK now depends on OEGrid TK. As result, CMake files might need to be modified to reflect this library dependency.

Java-specific changes

C#-specific changes

OEPlatform TK 2.1.5

  • Minor internal improvements have been made.

OESystem TK 2.1.5

  • The OEWriteSettings function now takes a 4th argument that specifies the name of an optimized parameter set for OpenEye‘s applications.

OEDepict TK 2.4.1

New features

Minor bug fixes

  • The minimum margin allowed for OE2DMolDisplayOptions has been reduced from 5.0% to 2.5%.
  • A duplicate SVG group warning is no longer thrown when transferring built-in SVG groups from one image to another.

OEDocking TK 1.3.7

  • Minor internal improvements have been made.

FastROCS TK 1.9.0

General notices

  • FastROCS TK no longer supports Nvidia Tesla, GEForce, and Quadro cards with a compute capability of less than 3.0. This includes all GPUs with Fermi architecture (e.g., Tesla C2050 and GEForce GTX 430). If you are running one of these older GPUs and cannot upgrade your hardware, do not update to this version of FastROCS TK.

New features

  • FastROCS TK is now supported for Java and C++. See the new documentation pages for Java and C++ for examples of toolkit usage and syntax.

    Note

    Java and C++ FastROCS TK API points are currently intended to be used only for toolkit functionality and not client-server functionality. The Python API should be used if client-server utilities are needed.

  • A new OEShapeDatabasePrep function has been added to provide the same functionality as the shapedatabaseprep example, except as a multi-threaded function. This should improve performance when prepping databases on systems with multiple cores. This function is non-deterministic, so molecules will not appear in the same order as the input file.

Major bug fixes

Documentation changes

  • The examples section of the documentation has been reorganized by type: Simple Examples, Alternative Starts, and Client-Server. All are viewable from the FastROCS Example Suite, which includes brief descriptions as well as links to the specific example pages. The example pages now provide more in-depth descriptions of usage and download links to the example code.
  • New examples have been added to the Alternative Starts section to demonstrate using FastROCS TK with alternative starts.
  • Documentation for both the Java and C++ versions of FastROCS TK has been added.

GraphemeTM TK 1.3.6

New features

New features (Preliminary API)

  • A new OERenderActiveSiteMaps function has been added that generates an interactive image of protein-ligand maps.

    Example of visualizing protein-ligand maps (PDB: 1br6)

Major bug fixes

  • When rendering ligand-protein maps with Grapheme TK, molecule surfaces are generated to depict the overall shape of the pocket. In the case of a large protein, this process can be both computationally expensive and can require a large amount of memory. In order to avoid these issues, only protein atoms that are no more than 20Å away from any ligand atoms are now considered and the resolution of the surface is reduced.
  • Rendering active site maps to an image with less than a 200.0 width or height parameter is now disabled to avoid generating very tiny, incomprehensible images.
  • When rendering ligand-protein maps (e.g., active site, unpaired), interactions between the ligand and more than 1 residue is now ignored and a warning is thrown.
  • Scaling of legend labels for large image sizes has been fixed. All Grapheme TK legends of ligand-protein depictions now have a consistent style.

Minor bug fixes

  • A problem that occurred when coloring hexagon glyphs representing DNA/RNA residues with more than 1 interaction type has been fixed.
  • An issue with depicting an interaction type in the legend when that interaction type is missing from the corresponding active site map has been fixed.
  • A small improvement has been made to help to generate residue layouts that are platform-independent.
  • A small improvement has been made to generate better residue layout: the space around the ligand for potential residue positions is now better sampled to allow to positioning residues close to edge of the canvas.
  • The text associated with the color stops when depicting color gradients vertically using the OEDrawColorGradient function is now right adjusted.

Documentation changes

GraphSim TK 2.4.1

New features

Minor bug fixes

  • The OEGetFingerPrintVersion function no longer throws an error for an invalid fingerprint type. It now throws a warning and returns zero.

Lexichem TK 2.6.6

  • Minor internal improvements have been made.

OEMedChem TK 1.0.7

New features

New features (Preliminary API)

  • The following preliminary APIs have been added to provide fast multithreaded generation of matched pair indices:
  • New OEReadMMPIndexMolecules and OEWriteMMPIndexMolecules functions have been implemented as preliminary APIs to support multithreaded and parallel matched pair index creation activities.
  • New function overloads to OEMatchedPairGetTransforms have been added to allow extracting transforms and related matched molecular pairs using one or more identified site(s) of substitution on a probe structure.
  • New free functions and overloads, OEGetMatchedMolecularPairs, have been added to allow extracting related matched molecular pairs using either a molecule ID or a probe structure with one or more identified site(s) of substitution.

API changes

Minor bug fixes

Python-specific changes

  • A new match pair index example has been added to demonstrate the creation of an MMP index via parallel processing using the Python multiprocessing module.

Documentation changes

  • The MCS fragment database examples and match pair index examples have been modified to separate the index creation step from the database query activities. See the list of currently available examples in the OEMedChem TK Examples section.

MolProp TK 2.5.1

  • Minor internal improvements have been made.

OEFF TK 2.0.3

New features

Minor bug fixes

  • Water is now treated as natural residue in the AMBER force field.

Documentation changes

Omega TK 2.8.0

New features (Preliminary API)

The following new preliminary APIs have been added for macrocycle functionality in the Omega TK:

The following new preliminary APIs have been added to enhance small molecule conformer generation functionality in the Omega TK:

New features

New additions and modifications have been made to existing APIs to enhance small molecule conformer generation functionality in Omega TK:

Major bug fixes

  • The SMILES string hash in the internal fragment library in OEOmega class has been updated to reflect the proper canonicalization of the fragments. This ensures that OEOmega will always find the fragment in the library if it exists.
  • The default search force field in OEOmega class has been changed to the newly introduced OEMMFFSheffieldFFType.MMFF94Smod_NOESTAT to generate more appropriate axial/equatorial conformers.
  • An issue that caused OEOmega class to not honor the parity bits on relative stereo marked rings has been fixed.

Minor bug fixes

  • An issue that caused OEOmega class to report the same warning about atoms multiple times has been fixed.

Documentation changes

The following C++ and Python examples have been added to demonstrate using the new preliminary APIs:

Quacpac TK 2.0.0

New features

  • A new OEGetReasonableTautomers function has been added that generates an ensemble of biologically relevant tautomers.
  • A new OEGetUniqueProtomer function has been added that generates a canonical representation of a molecule.
  • A new OEGetReasonableProtomers function has been added that generates an ensemble of reasonable protomers.
  • The handling of keto-enol tautomerization by default has been improved. The improvement extends to include imine-enamine tautomerization and related analogs. The keto and imine forms are favored except in the case of 1,3 diketones, where the internal hydrogen bond can be formed.
  • Controlling the interaction between tautomerization at tetrahedral stereochemistry has been improved. The method OETautomerOptions.SetSaveStereo can still be set to true in order to prevent any stereo atom from ever becoming an sp2 center. However, the behavior when the vale is set to false can now be controlled by passing a constant from the new OERacemicType namespace to the method OETautomerOptions.SetRacemicType. This functionality allows users to control which stereochemistry will be cleared on each generated tautomer.
  • Reasonable tautomers for many simple and complex pyridone analogs, pyrrole analogs, and pyridine analogs have been improved.
  • A new OEHypervalentNormalization function has been added to prepare molecules for tautomer enumeration (OEEnumerateTautomers).
  • The user control of the algorithms memory and CPU usage by separately exposing the number of tautomers generated and the number of tautomers returned by the methods OETautomerOptions.SetMaxTautomersGenerated and OETautomerOptions.SetMaxTautomersToReturn, respectively, has been improved.
  • The default parameters of the OETautomerOptions class have been updated to reflect the underlying improved tautomer generation algorithm.

Major bug fixes

  • Several issues that caused double bonds and hydrogens to be allowed to move too far through a series of sp3 centers have been fixed.
  • The default behavior of peptide tautomers and peptide stereochemistry has been improved.
  • Keto-enol tautomerization by default has been added to improve the canonical grouping of unique molecules by their representative tautomers.

Minor bug fixes

  • Aromatic bond count has been added to scoring to improve the handling of fused pyrrole and pyrazine chemistries when estimating low-energy tautomers.

Documentation changes

  • The tautomers theory chapter has been rewritten to reflect common usages of the tautomer library.

Shape TK 2.0.2

New features

Minor bug fixes

Java-specific changes

C#-specific changes

Documentation changes

Spicoli TK 1.4.5

New features

  • DNA/RNA ribbons are now available through the function OEMakeProteinRibbonSurface. When molecules containing DNA or RNA residues are passed to this function with a ribbon style of OERibbonType.Pretty, the resulting surface will contain ribbons for both protein and nucleic acid residues.

Major bug fixes

  • OEMakeAccessibleSurface no longer crashes when passed extremely large molecules (many hundreds of Angstroms across). The function now reduces the resolution of the surface, if necessary, to allow the surface to fit in the available memory.

Spruce TK 0.9.1

New features

Note

Spruce TK now requires a specific license. For information on Spruce TK licensing, please contact sales@eyesopen.com.

API changes

  • The constructor of the OEStructuralSuperposition class has been changed.
  • The low-level OEExtractBioUnitsFromRef and OEExtractBioUnitsFromPDBRemarks functions have been removed.

Major bug fixes

  • A bug that caused an issue with extracting biological assemblies has been fixed in the low-level biological assembly factory class.
  • Biological assembly extraction for large proteins has been improved for memory and CPU usage.

Minor bug fixes

  • Superposition methods for using OETrans instead of raw double pointers for rotation matrices and translation vectors have been improved.
  • A bug that caused an issue with extracting biological assemblies if they had a blank REMARK line in the PDB file has been fixed.

Szmap TK 1.4.4

  • Minor internal improvements have been made.

Szybki TK 2.0.3

Documentation changes

Zap TK 2.3.4

  • Minor internal improvements have been made.