OEChem TK 3.2.0¶

Fall 2021

New features¶

A new options class, OEUniMolecularRxnOptions, has been added to provide control arguments to the unimolecular reaction APIs OEUniMolecularRxn and OEGetUniMolecularRxnIter. The new options class adds support for enabling valence correction, as well as retrieval of altered atom information from the transform for post-processing of the transformed products.
Small molecule crystal structures are now writeable as CIF files, see OEWriteCIFFile for details.
Macromolecule files are now writeable as MMCIF files, see OEWriteCIFFile for details.
CIF and MMCIF files are now readable using oemolithread. This enables their use as OEMolDatabase as well.
The gemmi library used to read and parse MMCIF files has been updated to version 0.4.8.
New overloads of constructors for the OESimpleAppOptions and OERefInputAppOptions have been added to extend their abilities to be used when no sub-options are involved for an application or when custom file extensions are required.
New overloads of constructors for the OEFileStringParameter and OEOutFileStringParameter have been added that accepts file extensions as arguments instead of predefined constants.

An issue causing missing amide (peptide) bonds when reading protein structures from MMCIF files has been fixed.
An issue related to incorrect hydrogen atom stripping from stereocenters that are asymmetric by virtue of hydrogen isotope substitutions has been fixed. Isomeric SMILES for structures containing hydrogen isotopes on stereogenic centers will now have the atom/bond parity information correctly preserved. A minor consequence of the fix is that a small number of isomeric SMILES with explicit (non-isotopic) hydrogens may have SMILES differences due to conversion to a fully hydrogen-suppressed SMILES form.

An issue that could cause a crash during reads of OEFormat.OEB format files has been fixed. The read and write of OEGroupBase information in OEB files is now more rigorously checked for invalid objects. Existing corrupted OEB files will return an error status when read.
An issue that could cause overflow in the MDL molecule reader has been fixed.
An issue that caused return of an incorrect default value for OEPredicateParameter when it was set to a non-default value has been fixed.

The logic identifying aromatic nitrogen atoms has been updated in OEPlaceHydrogens, enabling improved hydrogen network optimization as well as the ability to properly distinguish and pick appropriate ligand tautomers in the protein binding site.

An issue causing the ligand density coverage (LaD) to change as a result of hydrogen placement in limited circumstances has been fixed.
An issue causing the operator bool on OEAtomMatchResidue to return true even if no valid data was provided to the constructor has been fixed.
An issue causing an OEDesignUnit attached as generic data on molecules to be lost during I/O has been fixed.
An issue causing proper backbone verification to fail for capping groups has been fixed.
An issue causing proper backbone verification to fail for non-standard amino acids has been fixed.
An issue causing histidine residues to be in the charged state (HIS+ or HIP) when it clearly was not favorable has been fixed.
The AspDyad mover rule in OEPlaceHydrogens has been refined.
The class OEPlaceHydrogensOptions has been modified to derive from OEOptions.

OEParameter classes that accept strings as input values now work with case-insensitive string inputs.
A new overload of OEMultiParameter.SetValue has been added that accepts oeifstream as input.

A new preliminary API class, OEKernelPLS, has been added that builds QSAR models from kernelized descriptors using partial least squares (PLS).