Szybki TK 1.8.7
================================================================================

New features
--------------------------------------------------------------------------------

* :oe:meth:`OESz::OESzybkiEnsembleResults::HasEntropy` has been added.
  This new method returns ``false`` whenever a requested entropy calculation fails.

* The default behavior of protein-ligand systems handled with MMFF or MMFF_AMBER 
  force fields has changed from :oe:const:`OESz::OEProteinElectrostatics::NoElectrostatics`
  to :oe:const:`OESz::OEProteinElectrostatics::ExactCoulomb`. In other words, 
  Coulomb protein-ligand interactions are included by default. 
  Beginning with this release, removing Coulomb terms from the protein-ligand 
  interactions requires using the ``OESz::OESzybkiOptions::SetProteinElectrostaticModel`` 
  method with the :oe:const:`OESz::OEProteinElectrostatics::NoElectrostatics` argument.

* Previously, the combination of IEFF force field and PB desolvation was restricted to
  using ESP charges calculated along with atomic multipoles. This restriction has been
  removed and IEFF for protein-ligand interaction can now be used together with the method
  ``OESz::OESzybkiOptions::SetUseCurrentCharges``. 

Major bug fixes
--------------------------------------------------------------------------------

* Return values of the methods ``OESz::OESzybkiOptions::SetExactVdWProteinLigand``
  and ``OESz::OESzybkiOptions::SetProteinFlexibilityRange`` have been changed
  from ``void`` to ``bool``. This change allows for verifying whether the options
  have been successfully set.

* When IEFF force field is selected, implying that molecular input files with 
  pre-calculated atomic multipoles have been used, the addition of explicit hydrogens is now disabled.
  In very rare cases (for example, tetracyanoquinodimethane ligand), perception of bonding might not
  correspond to the actual ab-initio pre-calculated structure, causing the IEFF 
  calculated energies to be incorrect. 

* Using external charges (see method ``OESz::OESzybkiOptions::SetUseCurrentCharges``)
  in the :oe:const:`OESz::OEForceFieldType::MMFF_AMBER` force field for protein-ligand Coulomb 
  interactions was failing unless the protein was eligible for correct AMBER charging. 
  This unnecessary condition has been removed: any atomic preassigned atomic charges 
  can now be used in combination with AMBER vdW protein-ligand interactions.

* Entropy calculation at very low temperatures is now handled properly:
  at 0 Kelvin, a value of zero is always returned.
  At temperatures close to 0 Kelvin, calculation is attempted and a warning is issued 
  if the calculation fails. 
  :oe:meth:`OESz::OESzybkiEnsembleResults::HasEntropy` (see above) might be used
  to query if the :ref:`OESzybkiResults <OESz_OESzybkiResults>` object contains valid
  calculated entropy.

* Using IEFF/PB potential for bound ligand optimization in torsion space when
  the ligand receptor is partially flexible did not always work.
  This issue has been fixed.

Minor bug fixes
--------------------------------------------------------------------------------

* When attempting to use Newton optimization and IEFF force field results, an
  error is now thrown with the  message "Second derivatives not available for IEFF ff".
