rdkit-devel — Developers information and discussion.

[Rdkit-devel] molecular representation underlying the postgresql cartridge for rdkit

[Peter S. <pet...@di...>]

Dear all,

I have a question regarding the underlying molecular representation used in the rdkit cartridge for postgresql.
I didn't dig yet too much into the underlying source code of the cartridge. Is it a normal rdkit like molecule used to store molecules?

What I want to make sure especially that it's not stored as smiles string!

Thanks in advance for any hint.

Peter

[Rdkit-devel] 2019.09.1 RDKit Release

[Greg L. <gre...@gm...>]

Dear all,

I'm pleased to announce that the next version of the RDKit - 2019.09 - is
released. The release notes are below.

The release files are on the github release page:
https://github.com/rdkit/rdkit/releases/tag/Release_2019_09_1

Binaries have been uploaded to anaconda.org (https://anaconda.org/rdkit).
The available conda binaries for this release are:
Linux 64bit: python 3.6, 3.7
Mac OS 64bit: python 3.6, 3.7
Windows 64bit: python 3.6, 3.7

Conda builds of the PostgreSQL cartridge are also available:
Linux 64bit: postgresql 9.6, 10, 11
Mac OS 64bit: postgresql 9.6, 10, 11

I believe that conda-forge will also switch to the new version in the near
future.

The online version of the documentation at rdkit.org (
http://rdkit.org/docs/index.html) has been updated.

Some things that will be finished over the next couple of days:
- The conda build scripts will be updated to reflect the new version
- The homebrew script

Thanks to everyone who submitted code, bug reports, and suggestions for
this release!

Please let me know if you find any problems with the release or have
suggestions for the next one, which is scheduled for March/April 2020.

Best Regards,
-greg

# Release_2019.09.1
(Changes relative to Release_2019.03.1)

## Important
- The atomic van der Waals radii used by the RDKit were corrected/updated
in #2154.
  This leads to different results when generating conformations, molecular
volumes,
  and molecular shapes.

## Backwards incompatible changes
- See the note about atomic van der Waals radii above.
- As part of the enhancements to the MolDraw2D class, we changed the type of
  DrawColour from a tuple to be an actual struct. We also added a 4th
element to
  capture alpha values. This should have no affect on Python code (the alpha
  value is optional when providing color tuples), but will require changes
to C++
  and Java/C# code that is using DrawColour.
- When reading Mol blocks, atoms with the symbol "R" are now converted into
  queries that match any atom when doing a substructure search (analogous
to "*"
  in SMARTS). The previous behavior was to only match other dummy atoms
- When loading SDF files using PandasTools.LoadSDF(), we now default to
  producing isomeric smiles in pandas tables.  To reproduce the original
  behavior, use isomericSmiles=False in the call to the function.
- The SMARTS generated by the RDKit no longer contains redundant wildcard
  queries. This means the SMARTS strings generated by this release will
generally
  be different from that in previous releases, although the results
produced by
  the queries should not change.
- The RGroupDecomposition code now removes Hs from output R groups by
default.
  To restore the old behavior create an RGroupDecompositionParameters
object and
  set removeHydrogensPostMatch to false.
- The default values for some of the new fingerprint generators have been
changed so
  that they more closely resemble the original fingerprinting code. In
  particular most fingerprinters no longer do count simulation by default
and
  the RDKit fingerprint now sets two bits per feature by default.
- The SMARTS generated for MCS results using the ringMatchesRingOnly or
  completeRingsOnly options now includes ring-membership queries.

## Highlights:
- The substructure matching code is now about 30% faster. This also
improves the
  speed of reaction matching and the FMCS code. (#2500)
- A minimal JavaScript wrapper has been added as part of the core release.
(#2444)
- It's now possible to get information about why molecule sanitization
failed. (#2587)
- A flexible new molecular hashing scheme has been added. (#2636)

## Acknowledgements:
Patricia Bento, Francois Berenger, Jason Biggs, David Cosgrove, Andrew
Dalke,
Thomas Duigou, Eloy Felix, Guillaume Godin, Lester Hedges, Anne Hersey,
Christoph Hillisch, Christopher Ing, Jan Holst Jensen, Gareth Jones, Eisuke
Kawashima, Brian Kelley, Alan Kerstjens, Karl Leswing, Pat Lorton, John
Mayfield, Mike Mazanetz, Dan Nealschneider, Noel O'Boyle, Stephen Roughley,
Roger Sayle, Ricardo Rodriguez Schmidt, Paula Schmiel, Peter St. John,
Marvin
Steijaert, Matt Swain, Amol Thakkar Paolo Tosco, Yi-Shu Tu, Ricardo
Vianello,
Marc Wittke, '7FeiW', 'c56pony', 'sirbiscuit'


## Bug Fixes:
  - MCS returning partial rings with completeRingsOnly=True
 (github issue #945 from greglandrum)
  - Alternating canonical SMILES for fused ring with N
 (github issue #1028 from greglandrum)
  - Atom index out of range error
 (github issue #1868 from A-Thakkar)
  - Incorrect cis/trans stereo symbol for conjugated ring
 (github issue #2023 from baoilleach)
  - Hardcoded max length of SMARTs string cut of input query for FragCatlog
 (github issue #2163 from 7FeiW)
  - VSA_EState {1, ..., 10} calculated by rdkit doesn't seem correct.
 (github issue #2372 from c56pony)
  - MolStandardize: FragmentRemover should not sanitize fragments
 (github issue #2411 from greglandrum)
  - MolStandardize: combinatorial explosion in Normalizer
 (github issue #2414 from greglandrum)
  - MCS code doesn't return envelope MCS if CompleteRingsOnly is true
 (github issue #2420 from greglandrum)
  - RemoveHs() does not remove all hydrogens.
 (github issue #2422 from paulaju)
  - Incorrect assignment of explicit Hs to Al+3 read from mol block
 (github issue #2423 from greglandrum)
  - Cannot set maxProducts > 1000 in RunReactants
 (github issue #2427 from tduigou)
  - Chem.MolStandardize.standardize.Standardizer drops molecular properties
 (github pull #2431 from lilleswing)
  - Canon::rankMolAtoms results in crossed double bonds in rings
 (github issue #2437 from greglandrum)
  - make boost::iostreams optional
 (github pull #2440 from greglandrum)
  - Fix/rgroup sdf isotope
 (github pull #2449 from bp-kelley)
  - Uncharger incorrectly removing charge from boron anions
 (github issue #2452 from greglandrum)
  - Add java builds to azure devops
 (github pull #2460 from greglandrum)
  - Cart fixes
 (github pull #2462 from jones-gareth)
  - Negative atom map values cause depickling to fail
 (github issue #2465 from greglandrum)
  - Deserialization failures crash Java wrapper
 (github issue #2466 from greglandrum)
  - rdkit.six fix and cleanup
 (github pull #2469 from rvianello)
  - dummy atom queries are flagged as complex
 (github issue #2471 from greglandrum)
  - 3D structure display broken in jupyter notebook
 (github issue #2473 from greglandrum)
  - Inconsistent defaults for nonBondedThresh in MMFF optimization
 (github issue #2475 from greglandrum)
  - Fix/rgroup multiple labels
 (github pull #2481 from bp-kelley)
  - 2D Depiction clipped atom highlighting
 (github issue #2486 from DavidACosgrove)
  - BRICSBuild now passes scrambleReagents to children
 (github pull #2488 from greglandrum)
  - Pattern Fingerprint Issues
 (github issue #2501 from jones-gareth)
  - CMake Error: Wrap directories being used when python build is turned off
 (github issue #2516 from jasondbiggs)
  - - fixes ResonanceMolSupplier bug in perceiving conjugated groups
 (github pull #2517 from ptosco)
  - Fix/mmff threadsafety issues
 (github pull #2518 from bp-kelley)
  - update expected SVG output in cartridge tests
 (github pull #2520 from greglandrum)
  - fix to SDWriter docs
 (github pull #2521 from pstjohn)
  - Fix the azure pipelines builds
 (github pull #2522 from greglandrum)
  - Code cleanups from PVS/Studio
 (github pull #2531 from greglandrum)
  - getAtomNeighbors() and getAtomBonds() not in SWIG wrappers.
 (github issue #2532 from greglandrum)
  - Default sanitizerxn doesn't aromatize if possible
 (github issue #2547 from bp-kelley)
  - Add RDKIT_FILEPARSERS_EXPORT to finishMolProcessing
 (github pull #2551 from d-b-w)
  - Chem.rdFMCS.FindMCS hangs for certain ligand pairs
 (github issue #2581 from lohedges)
  - fix the inclusion path for the targets file (#2584)
 (github pull #2589 from rvianello)
  - Fix inocuous typo/bug in Dative bond matching
 (github pull #2593 from ricrogz)
  - E/Z and CIS/TRANS stereo bonds are incorrectly matched
 (github pull #2596 from ricrogz)
  - Uncharger ignores dications
 (github issue #2602 from greglandrum)
  - Possible Garbage Collection Bug in Pharmacophore Generation
 (github issue #2603 from cing)
  - Uncharger incorrectly neutralizes cations when non-neutralizable anions
are present.
 (github issue #2605 from greglandrum)
  - Bad valence corrections on Pb, Sn
 (github issue #2606 from greglandrum)
  - Pb and Sn should support valence 2
 (github issue #2607 from greglandrum)
  - Uncharger incorrectly modifying a zwitterion
 (github issue #2610 from greglandrum)
  - CanonicalRankAtomsInFragment breakTies doesn't
 (github issue #2611 from bp-kelley)
  - Pattern fingerprint failing substructure condition in very large
molecules
 (github issue #2614 from greglandrum)
  - Memory leak with Chem.Atom() constructor
 (github issue #2639 from AlanKerstjens)
  - Fixes: Atoms in non-standard valences not being properly written to mol
blocks
 (github pull #2646 from greglandrum)
  - C++ MCS code returns a null MCS between methylcyclopentane and
methylcyclohexane
 (github issue #2662 from ptosco)
  - CXSMILES writer has error if mol comes from v3000 molfile
 (github issue #2666 from d-b-w)
  - MolToCXSmiles generates error for empty molecule
 (github issue #2667 from greglandrum)
  - fix a problem with normalize, ringinfo, and fragments
 (github pull #2685 from greglandrum)
  - Error when a squiggle bond is in an aromatic ring
 (github issue #2695 from greglandrum)
  - Cannot combine multiple range queries on a single atom.
 (github issue #2709 from greglandrum)
  - setBondStereoFromDirections() returning incorrect results.
 (github issue #2712 from greglandrum)
  - update supplier documentation to reflect python 3 iterator syntax
 (github pull #2719 from greglandrum)
  - removeHs messing up double bond stereo in partially sanitized molecules
 (github issue #2721 from greglandrum)
  - seg fault in ReactionRunner
 (github issue #2722 from greglandrum)
  - Intermittent test failures for JavaDistanceGeometryTests
 (github issue #2727 from greglandrum)
  - Fixes a bug in TorsionConstraint
 (github pull #2732 from ptosco)
  - Apply fix for #1592 to _MolsToGridSVG
 (github pull #2737 from yishutu)

## New Features and Enhancements:
  - Added rankAtoms to ROMol wrapper and added Java test case
 (github pull #1540 from sroughley)
  - Use van der Waals radii from blue obelisk
 (github pull #2154 from UnixJunkie)
  - add generateDepictionMatching2DStructure() to SWIG wrappers
 (github issue #2239 from greglandrum)
  - Added OptimizeMoleculeConfs with pre-generated force-field
 (github pull #2401 from ptosco)
  - FreeSASA improvements
 (github pull #2402 from ptosco)
  - Speed up symmetrizeSSSR
 (github issue #2403 from d-b-w)
  - Trim whitespace from mol fragment SMARTS and check SMARTS presence
 (github pull #2406 from ricrogz)
  - Run clang-tidy over the entire codebase
 (github pull #2408 from greglandrum)
  - Enable Azure Pipelines builds for CI
 (github pull #2409 from ricrogz)
  - Add RDProps interface to Conformers
 (github issue #2441 from greglandrum)
  - Add minimal JavaScript wrapper
 (github pull #2444 from greglandrum)
  - Fixes annoying warnings on MSVC
 (github pull #2454 from ptosco)
  - add prepareMolsBeforeDrawing option for drawMols
 (github pull #2455 from greglandrum)
  - computeGasteigerCharges quality of life improvement for python api
 (github issue #2480 from bp-kelley)
  - Preserve bond direction in fragmentOnBonds
 (github pull #2484 from greglandrum)
  - SanitizeRxn code and docstring cleanup
 (github pull #2491 from greglandrum)
  - Support XYZ format for output
 (github pull #2498 from e-kwsm)
  - vf2 optimisations
 (github pull #2500 from johnmay)
  - Python wrap enhanced stereo setters
 (github pull #2509 from d-b-w)
  - Fix the azure pipelines builds
 (github pull #2522 from greglandrum)
  - add a script for benchmarking fingerprint screenout and substructure
performance
 (github pull #2523 from greglandrum)
  - make "R" in CTABs an AtomNull query
 (github pull #2528 from greglandrum)
  - Expose SDF loading options to LoadSDF
 (github pull #2534 from bp-kelley)
  - Remove unused ctest: testCanon
 (github pull #2541 from ricrogz)
  - Update maeparser and coordgen versions
 (github pull #2542 from ricrogz)
  - Improved handling of bond stereo in reactions
 (github pull #2553 from ricrogz)
  - Code simplification for fingerprints to np array
 (github pull #2557 from ChrisHill8)
  - Integrate Unique Ring Families from RingDecomposerLib
 (github pull #2558 from greglandrum)
  - Allow providing a bounds matrix to EmbedMol
 (github pull #2560 from greglandrum)
  - Enable SimilarityMaps in C++
 (github pull #2562 from greglandrum)
  - Do not run UnitTestMCS.py::TestTimeout in debug builds
 (github pull #2569 from ricrogz)
  - Expose more drawing methods to Python
 (github issue #2571 from greglandrum)
  - Allow Point2D to be constructed from Point3D
 (github pull #2572 from greglandrum)
  - Allows dative bonds to be drawn
 (github pull #2573 from greglandrum)
  - Allow identification of chemistry problems
 (github pull #2587 from greglandrum)
  - Adds MolFragmentToSmarts to generate smarts for a subset of a Molecule
 (github pull #2594 from d-b-w)
  - Removal of redundant wildcards in SMARTS (Null Atom/Bond Query
combination)
 (github pull #2595 from ricrogz)
  - Support range-based charge queries from SMARTS
 (github issue #2604 from greglandrum)
  - Keep PDB info from Maestro files if available
 (github pull #2619 from lorton)
  - optimization of the MolStandardize code
 (github pull #2621 from greglandrum)
  - Assign stereo bond labels in molecules read from SMARTS
 (github pull #2623 from ricrogz)
  - Automatically load the ipython extensions running in Jupyter
 (github pull #2626 from bp-kelley)
  - draw zero-order bonds
 (github pull #2630 from greglandrum)
  - Updated cartridge documentation
 (github pull #2635 from msteijaert)
  - Add new mol hashing code
 (github pull #2636 from greglandrum)
  - emolecules link updated
 (github pull #2638 from marcwittke)
  - Update maeparser to 1.2.1 and coorgen to 1.3.1
 (github pull #2652 from ricrogz)
  - Get numpy include dir programmatically
 (github pull #2653 from sirbiscuit)
 - Fix long columns pandas
 (github pull #2655 from sirbiscuit)
  - Added AtomComparator.AtomCompareAnyHeavyAtom and test cases to FMCS code
 (github pull #2656 from sroughley)
  - The C++ MCS code generates ambiguous SMARTS strings
 (github issue #2663 from ptosco)
  - add bond-selector info to SVGs
 (github pull #2664 from greglandrum)
  - support writing CXSMILES from the cartridge
 (github issue #2668 from greglandrum)
  - support the new hashing code in the cartridge
 (github pull #2671 from greglandrum)
  - Adds additional capabilities to the minimal JS wrapper
 (github pull #2676 from greglandrum)
  - Add MolHash to Java Wrappers
 (github issue #2677 from sroughley)
  - A bunch of changes to the new fingerprinter code
 (github pull #2679 from greglandrum)
  - Add viewBox to default SVG output
 (github issue #2680 from bp-kelley)
  - Allow Java to see RGroup labels in the std::map wrapper.
 (github pull #2681 from bp-kelley)
  - Update maeparser to v1.2.2.
 (github pull #2682 from ricrogz)
  - Update coordgen to v1.3.2
 (github pull #2686 from ricrogz)
  - Add a drawOptions object to IPythonConsole
 (github pull #2691 from greglandrum)
  - Make StructureGroups editable from Python
 (github pull #2692 from greglandrum)
  - Update documentation
 (github pull #2697 from greglandrum)
  - Make removeHydrogensPostMatch=true the default for RGD
 (github pull #2701 from greglandrum)
  - Eat our own dogfood, Clone is deprecated so use copy
 (github pull #2711 from bp-kelley)
  - The MCS smartsString may still be ambiguous
 (github issue #2714 from ptosco)
  - Add threaded runner for the filter catalog
 (github pull #2718 from bp-kelley)
  - Add Leader picker implementation
 (github pull #2724 from greglandrum)
  - Add consideration of ring fusion to the MCS algorithm
 (github pull #2731 from ptosco)


## Deprecated code (to be removed in a future release):

- The old MolHash code should be considered deprecated. This release
introduces
  a more flexible alternative. Specifically the following pieces will be
removed in a future release:
  - The python functionality
`rdkit.Chem.rdMolHash.GenerateMoleculeHashString()`
  - The C++ functionality directly present in the header file
`GraphMol/MolHash/MolHash.h`

[Rdkit-devel] Beta of the 2019.09 release available

[Greg L. <gre...@gm...>]

Dear all,

The beta of the 2019.09 RDKit release has been tagged in github:
<https://github.com/rdkit/rdkit/releases/tag/Release_2019_03_1b1>
https://github.com/rdkit/rdkit/releases/tag/Release_2019_09_1b1

There are a couple more bug fixes and hopefully one more feature expected
before the actual release, but I wanted to go ahead and get the beta out
there.

I've done conda builds for Python 3.6 and 3.7 for Mac, and Linux and I'm
working on the Windows builds now. These all use the beta label so that
they do not install by default; you'll need to run "conda install" as
follows:

conda install -c rdkit/label/beta rdkit

Be sure to confirm that it's installing the right version when you are
prompted (if there's no build available, it will pick the current
production release instead).

The relevant section of the release notes is below, or you can see a nicely
formatted version here:
<https://github.com/rdkit/rdkit/releases/tag/Release_2019_03_1b1>
https://github.com/rdkit/rdkit/releases/tag/Release_2019_09_1b1

As usual, if you have time to try out the new release I would love
feedback. If nothing major comes up, I plan to do the actual release
towards the end of next week.

Best,
-greg

# Release_2019.09.1
(Changes relative to Release_2019.03.1)

## Important
- The atomic van der Waals radii used by the RDKit were corrected/updated
in #2154.
  This leads to different results when generating conformations, molecular
volumes,
  and molecular shapes.

## Backwards incompatible changes
- See the note about atomic van der Waals radii above.
- As part of the enhancements to the MolDraw2D class, we changed the type of
  DrawColour from a tuple to be an actual struct. We also added a 4th
element to
  capture alpha values. This should have no affect on Python code (the alpha
  value is optional when providing color tuples), but will require changes
to C++
  and Java/C# code that is using DrawColour.
- When reading Mol blocks, atoms with the symbol "R" are now converted into
  queries that match any atom when doing a substructure search (analogous
to "*"
  in SMARTS). The previous behavior was to only match other dummy atoms
- When loading SDF files using PandasTools.LoadSDF(), we now default to
  producing isomeric smiles in pandas tables.  To reproduce the original
  behavior, use isomericSmiles=False in the call to the function.
- The SMARTS generated by the RDKit no longer contains redundant wildcard
  queries. This means the SMARTS strings generated by this release will
generally
  be different from that in previous releases, although the results
produced by
  the queries should not change.
- The RGroupDecomposition code now removes Hs from output R groups by
default.
  To restore the old behavior create an RGroupDecompositionParameters
object and
  set removeHydrogensPostMatch to false.
- The default values for some of the new fingerprint generators have been
changed so
  that they more closely resemble the original fingerprinting code. In
  particular most fingerprinters no longer do count simulation by default
and
  the RDKit fingerprint now sets two bits per feature by default.
- The SMARTS generated for MCS results using the ringMatchesRingOnly or
  completeRingsOnly options now includes ring-membership queries.

## Highlights:

## Acknowledgements:
Patricia Bento, Francois Berenger, Jason Biggs, David Cosgrove, Andrew
Dalke,
Thomas Duigou, Eloy Felix, Guillaume Godin, Lester Hedges, Anne Hersey,
Christoph Hillisch, Christopher Ing, Jan Holst Jensen, Gareth Jones, Eisuke
Kawashima, Brian Kelley, Alan Kerstjens, Karl Leswing, Pat Lorton, John
Mayfield, Mike Mazanetz, Dan Nealschneider, Noel O'Boyle, Stephen Roughley,
Roger Sayle, Ricardo Rodriguez Schmidt, Paula Schmiel, Peter St. John, Matt
Swain, Amol Thakkar Paolo Tosco, Ricardo Vianello, Marc Wittke,
'7FeiW', 'c56pony', 'msteijaert', 'sirbiscuit'


## Bug Fixes:
  - MCS returning partial rings with completeRingsOnly=True
 (github issue #945 from greglandrum)
  - Alternating canonical SMILES for fused ring with N
 (github issue #1028 from greglandrum)
  - Atom index out of range error
 (github issue #1868 from A-Thakkar)
  - Incorrect cis/trans stereo symbol for conjugated ring
 (github issue #2023 from baoilleach)
  - Hardcoded max length of SMARTs string cut of input query for FragCatlog
 (github issue #2163 from 7FeiW)
  - VSA_EState {1, ..., 10} calculated by rdkit doesn't seem correct.
 (github issue #2372 from c56pony)
  - MolStandardize: FragmentRemover should not sanitize fragments
 (github issue #2411 from greglandrum)
  - MolStandardize: combinatorial explosion in Normalizer
 (github issue #2414 from greglandrum)
  - MCS code doesn't return envelope MCS if CompleteRingsOnly is true
 (github issue #2420 from greglandrum)
  - RemoveHs() does not remove all hydrogens.
 (github issue #2422 from paulaju)
  - Incorrect assignment of explicit Hs to Al+3 read from mol block
 (github issue #2423 from greglandrum)
  - Cannot set maxProducts > 1000 in RunReactants
 (github issue #2427 from tduigou)
  - Chem.MolStandardize.standardize.Standardizer drops molecular properties
 (github pull #2431 from lilleswing)
  - Canon::rankMolAtoms results in crossed double bonds in rings
 (github issue #2437 from greglandrum)
  - make boost::iostreams optional
 (github pull #2440 from greglandrum)
  - Fix/rgroup sdf isotope
 (github pull #2449 from bp-kelley)
  - Uncharger incorrectly removing charge from boron anions
 (github issue #2452 from greglandrum)
  - Add java builds to azure devops
 (github pull #2460 from greglandrum)
  - Cart fixes
 (github pull #2462 from jones-gareth)
  - Negative atom map values cause depickling to fail
 (github issue #2465 from greglandrum)
  - Deserialization failures crash Java wrapper
 (github issue #2466 from greglandrum)
  - rdkit.six fix and cleanup
 (github pull #2469 from rvianello)
  - dummy atom queries are flagged as complex
 (github issue #2471 from greglandrum)
  - 3D structure display broken in jupyter notebook
 (github issue #2473 from greglandrum)
  - Inconsistent defaults for nonBondedThresh in MMFF optimization
 (github issue #2475 from greglandrum)
  - Fix/rgroup multiple labels
 (github pull #2481 from bp-kelley)
  - 2D Depiction clipped atom highlighting
 (github issue #2486 from DavidACosgrove)
  - BRICSBuild now passes scrambleReagents to children
 (github pull #2488 from greglandrum)
  - Pattern Fingerprint Issues
 (github issue #2501 from jones-gareth)
  - CMake Error: Wrap directories being used when python build is turned off
 (github issue #2516 from jasondbiggs)
  - - fixes ResonanceMolSupplier bug in perceiving conjugated groups
 (github pull #2517 from ptosco)
  - Fix/mmff threadsafety issues
 (github pull #2518 from bp-kelley)
  - update expected SVG output in cartridge tests
 (github pull #2520 from greglandrum)
  - fix to SDWriter docs
 (github pull #2521 from pstjohn)
  - Fix the azure pipelines builds
 (github pull #2522 from greglandrum)
  - Code cleanups from PVS/Studio
 (github pull #2531 from greglandrum)
  - getAtomNeighbors() and getAtomBonds() not in SWIG wrappers.
 (github issue #2532 from greglandrum)
  - Default sanitizerxn doesn't aromatize if possible
 (github issue #2547 from bp-kelley)
  - Add RDKIT_FILEPARSERS_EXPORT to finishMolProcessing
 (github pull #2551 from d-b-w)
  - Chem.rdFMCS.FindMCS hangs for certain ligand pairs
 (github issue #2581 from lohedges)
  - fix the inclusion path for the targets file (#2584)
 (github pull #2589 from rvianello)
  - Fix inocuous typo/bug in Dative bond matching
 (github pull #2593 from ricrogz)
  - E/Z and CIS/TRANS stereo bonds are incorrectly matched
 (github pull #2596 from ricrogz)
  - Uncharger ignores dications
 (github issue #2602 from greglandrum)
  - Possible Garbage Collection Bug in Pharmacophore Generation
 (github issue #2603 from cing)
  - Uncharger incorrectly neutralizes cations when non-neutralizable anions
are present.
 (github issue #2605 from greglandrum)
  - Bad valence corrections on Pb, Sn
 (github issue #2606 from greglandrum)
  - Pb and Sn should support valence 2
 (github issue #2607 from greglandrum)
  - Uncharger incorrectly modifying a zwitterion
 (github issue #2610 from greglandrum)
  - CanonicalRankAtomsInFragment breakTies doesn't
 (github issue #2611 from bp-kelley)
  - Pattern fingerprint failing substructure condition in very large
molecules
 (github issue #2614 from greglandrum)
  - Memory leak with Chem.Atom() constructor
 (github issue #2639 from AlanKerstjens)
  - Fixes: Atoms in non-standard valences not being properly written to mol
blocks
 (github pull #2646 from greglandrum)
  - C++ MCS code returns a null MCS between methylcyclopentane and
methylcyclohexane
 (github issue #2662 from ptosco)
  - CXSMILES writer has error if mol comes from v3000 molfile
 (github issue #2666 from d-b-w)
  - MolToCXSmiles generates error for empty molecule
 (github issue #2667 from greglandrum)
  - fix a problem with normalize, ringinfo, and fragments
 (github pull #2685 from greglandrum)
  - Error when a squiggle bond is in an aromatic ring
 (github issue #2695 from greglandrum)
  - Cannot combine multiple range queries on a single atom.
 (github issue #2709 from greglandrum)
  - setBondStereoFromDirections() returning incorrect results.
 (github issue #2712 from greglandrum)
  - update supplier documentation to reflect python 3 iterator syntax
 (github pull #2719 from greglandrum)
  - removeHs messing up double bond stereo in partially sanitized molecules
 (github issue #2721 from greglandrum)
  - seg fault in ReactionRunner
 (github issue #2722 from greglandrum)

## New Features and Enhancements:
  - Added rankAtoms to ROMol wrapper and added Java test case
 (github pull #1540 from sroughley)
  - Use van der Waals radii from blue obelisk
 (github pull #2154 from UnixJunkie)
  - add generateDepictionMatching2DStructure() to SWIG wrappers
 (github issue #2239 from greglandrum)
  - Added OptimizeMoleculeConfs with pre-generated force-field
 (github pull #2401 from ptosco)
  - FreeSASA improvements
 (github pull #2402 from ptosco)
  - Speed up symmetrizeSSSR
 (github issue #2403 from d-b-w)
  - Trim whitespace from mol fragment SMARTS and check SMARTS presence
 (github pull #2406 from ricrogz)
  - Run clang-tidy over the entire codebase
 (github pull #2408 from greglandrum)
  - Enable Azure Pipelines builds for CI
 (github pull #2409 from ricrogz)
  - Add RDProps interface to Conformers
 (github issue #2441 from greglandrum)
  - Add minimal JavaScript wrapper
 (github pull #2444 from greglandrum)
  - Fixes annoying warnings on MSVC
 (github pull #2454 from ptosco)
  - add prepareMolsBeforeDrawing option for drawMols
 (github pull #2455 from greglandrum)
  - computeGasteigerCharges quality of life improvement for python api
 (github issue #2480 from bp-kelley)
  - Preserve bond direction in fragmentOnBonds
 (github pull #2484 from greglandrum)
  - SanitizeRxn code and docstring cleanup
 (github pull #2491 from greglandrum)
  - Support XYZ format for output
 (github pull #2498 from e-kwsm)
  - vf2 optimisations
 (github pull #2500 from johnmay)
  - Python wrap enhanced stereo setters
 (github pull #2509 from d-b-w)
  - Fix the azure pipelines builds
 (github pull #2522 from greglandrum)
  - add a script for benchmarking fingerprint screenout and substructure
performance
 (github pull #2523 from greglandrum)
  - make "R" in CTABs an AtomNull query
 (github pull #2528 from greglandrum)
  - Expose SDF loading options to LoadSDF
 (github pull #2534 from bp-kelley)
  - Remove unused ctest: testCanon
 (github pull #2541 from ricrogz)
  - Update maeparser and coordgen versions
 (github pull #2542 from ricrogz)
  - Improved handling of bond stereo in reactions
 (github pull #2553 from ricrogz)
  - Code simplification for fingerprints to np array
 (github pull #2557 from ChrisHill8)
  - Integrate Unique Ring Families from RingDecomposerLib
 (github pull #2558 from greglandrum)
  - Allow providing a bounds matrix to EmbedMol
 (github pull #2560 from greglandrum)
  - Enable SimilarityMaps in C++
 (github pull #2562 from greglandrum)
  - Do not run UnitTestMCS.py::TestTimeout in debug builds
 (github pull #2569 from ricrogz)
  - Expose more drawing methods to Python
 (github issue #2571 from greglandrum)
  - Allow Point2D to be constructed from Point3D
 (github pull #2572 from greglandrum)
  - Allows dative bonds to be drawn
 (github pull #2573 from greglandrum)
  - Allow identification of chemistry problems
 (github pull #2587 from greglandrum)
  - Adds MolFragmentToSmarts to generate smarts for a subset of a Molecule
 (github pull #2594 from d-b-w)
  - Removal of redundant wildcards in SMARTS (Null Atom/Bond Query
combination)
 (github pull #2595 from ricrogz)
  - Support range-based charge queries from SMARTS
 (github issue #2604 from greglandrum)
  - Keep PDB info from Maestro files if available
 (github pull #2619 from lorton)
  - optimization of the MolStandardize code
 (github pull #2621 from greglandrum)
  - Assign stereo bond labels in molecules read from SMARTS
 (github pull #2623 from ricrogz)
  - Automatically load the ipython extensions running in Jupyter
 (github pull #2626 from bp-kelley)
  - draw zero-order bonds
 (github pull #2630 from greglandrum)
  - Updated cartridge documentation
 (github pull #2635 from msteijaert)
  - Add new mol hashing code
 (github pull #2636 from greglandrum)
  - emolecules link updated
 (github pull #2638 from marcwittke)
  - Update maeparser to 1.2.1 and coorgen to 1.3.1
 (github pull #2652 from ricrogz)
  - Get numpy include dir programmatically
 (github pull #2653 from sirbiscuit)
 - Fix long columns pandas
 (github pull #2655 from sirbiscuit)
  - Added AtomComparator.AtomCompareAnyHeavyAtom and test cases to FMCS code
 (github pull #2656 from sroughley)
  - The C++ MCS code generates ambiguous SMARTS strings
 (github issue #2663 from ptosco)
  - add bond-selector info to SVGs
 (github pull #2664 from greglandrum)
  - support writing CXSMILES from the cartridge
 (github issue #2668 from greglandrum)
  - support the new hashing code in the cartridge
 (github pull #2671 from greglandrum)
  - Adds additional capabilities to the minimal JS wrapper
 (github pull #2676 from greglandrum)
  - Add MolHash to Java Wrappers
 (github issue #2677 from sroughley)
  - A bunch of changes to the new fingerprinter code
 (github pull #2679 from greglandrum)
  - Add viewBox to default SVG output
 (github issue #2680 from bp-kelley)
  - Allow Java to see RGroup labels in the std::map wrapper.
 (github pull #2681 from bp-kelley)
  - Update maeparser to v1.2.2.
 (github pull #2682 from ricrogz)
  - Update coordgen to v1.3.2
 (github pull #2686 from ricrogz)
  - Add a drawOptions object to IPythonConsole
 (github pull #2691 from greglandrum)
  - [WIP] Make StructureGroups editable from Python
 (github pull #2692 from greglandrum)
  - Update documentation
 (github pull #2697 from greglandrum)
  - Make removeHydrogensPostMatch=true the default for RGD
 (github pull #2701 from greglandrum)
  - Eat our own dogfood, Clone is deprecated so use copy
 (github pull #2711 from bp-kelley)
  - The MCS smartsString may still be ambiguous
 (github issue #2714 from ptosco)

## Deprecated code (to be removed in a future release):

- The old MolHash code should be considered deprecated. This release
introduces
  a more flexible alternative. Specifically the following pieces will be
removed in a future release:
  - The python functionality
`rdkit.Chem.rdMolHash.GenerateMoleculeHashString()`
  - The C++ functionality directly present in the header file
`GraphMol/MolHash/MolHash.h`

[Rdkit-devel] CI Update

[Greg L. <gre...@gm...>]

Dear all,

I just realized that I haven't really said anything here about the changes
that we've been making to the CI (continuous integration) system that we're
using for the RDKit.

In case you have no idea what I'm talking about or use CI in a different
way: every pull request or code change that's done on the RDKit github repo
triggers a number of automated processes that build and test the the code
and then report back whether or not they succeeded.

Up until recently we had been using two different systems for this:
TravisCI (https://travis-ci.com/ for Linux and Mac builds) and Appveyor (
https://www.appveyor.com/ for Windows builds). These companies provide a
great service to the open-source community by providing free access to
their systems for open-source projects. This is just a fantastic thing. The
Appveyor and TravisCI builds have really helped us ensure that the RDKit is
still passing all of the tests what we've written.

In the reasonably recent past (to be honest, I'm not quite sure when),
Microsoft started providing a similar service, Azure DevOps (
https://azure.microsoft.com/en-us/services/devops/), that has the very nice
property that it supports Linux, Windows, and Mac. Back in April Ricardo
(@ricrogz for those of you watching github) figured out how to make the
RDKit builds work with Azure DevOps and submitted the PR to make it all
work. Since then every PR and code change pushed is also sent to Azure
DevOps where the code is built and tested on all three operating systems.
I'm really happy to have everything in one place (particularly since the
build scripts and configuration are pretty easy to modify now that Ricardo
got the initial version working).

We just extended the Azure DevOps build matrix so that it automatically
builds and tests the Java wrappers on Linux and the Mac, which is going to
be a big help for me when it comes to release time. We'll be adding a
Windows version of that as well (testing both the Java and the C#
wrappers). If things go well we can also add tests for the cartridge builds
in the near future. This is great stuff!

Given that we now have good Windows CI through the Azure services and that
the Appveyor RDKit builds have always been kind of second-rate (this isn't
an Appveyor problem: the RDKit is a bit too big for what is available as a
free service from Appveyor), I just disabled the Appveyor integration.

-greg

Re: [Rdkit-devel] GSoC-2019 RDKit-Neo4J project

[Greg L. <gre...@gm...>]

Hi Evgeny,

Thanks for introducing yourself! I'm really working forward to working with
you on this project and seeing what we can do in order to allow people to
use the RDKit together with Neo4j.

I've created a github repo for this project here:
https://github.com/rdkit/neo4j-rdkit
and I encourage anyone who's interested to keep an eye on that.

Best,
-greg



On Thu, May 9, 2019 at 5:27 PM Evgeniy Sorokin <
evg...@gm...> wrote:

> Hello everyone,
> I am Evgeny Sorokin from Russia and I study in Innopolis University (3rd
> year). It is a new city and a new university that appeared less than 6
> years ago. I am mainly Java/Python developer with additional experience in
> finance/blockchain/ml projects.
> This is my second time I participate in GSoC, previously I worked with
> Tomas Pluskal on MSDK/MZMine projects.
>
> During this summer will be working on RDKit-Neo4J project
> <https://docs.google.com/document/d/1rXUbY3di_YSW1LWTraa3cJEapRtEOrcBc1otZV56FMc/edit#> with
> Greg Landrum, Christian Pilger and Stefan Armbruster (neo4j core
> developer).
> The project will be focused on development of extension for neo4j graph
> database for querying knowledge graphs storing molecular and chemical
> information. This RDKit functionality is already implemented and works with
> Postgres db, the same result is expected to be accomplished during this
> summer for neo4j.
> Best regards, Evgeny Sorokin
> --- @evgerher in telegram, github
> _______________________________________________
> Rdkit-devel mailing list
> Rdk...@li...
> https://lists.sourceforge.net/lists/listinfo/rdkit-devel
>

[Rdkit-devel] GSoC-2019 RDKit-Neo4J project

[Evgeniy S. <evg...@gm...>]

Hello everyone,
I am Evgeny Sorokin from Russia and I study in Innopolis University (3rd
year). It is a new city and a new university that appeared less than 6
years ago. I am mainly Java/Python developer with additional experience in
finance/blockchain/ml projects.
This is my second time I participate in GSoC, previously I worked with
Tomas Pluskal on MSDK/MZMine projects.

During this summer will be working on RDKit-Neo4J project
<https://docs.google.com/document/d/1rXUbY3di_YSW1LWTraa3cJEapRtEOrcBc1otZV56FMc/edit#>
with
Greg Landrum, Christian Pilger and Stefan Armbruster (neo4j core developer).
The project will be focused on development of extension for neo4j graph
database for querying knowledge graphs storing molecular and chemical
information. This RDKit functionality is already implemented and works with
Postgres db, the same result is expected to be accomplished during this
summer for neo4j.
Best regards, Evgeny Sorokin
--- @evgerher in telegram, github

[Rdkit-devel] GSoC 2019 project introduction

[Naruki Y. <nar...@gm...>]

Hi everyone,

I am going to work on the GSoC 2019 project "More Improved Coordinate
Generation for Open Babel: Combining Fragment-based Method and
Distance Geometry"
https://summerofcode.withgoogle.com/projects/#5547940045651968
Although this is a project for Open Babel, I'd like to introduce to
developers of RDKit too.

The goal of this project is to improve the current coordinate
generation method by incorporating the distance geometry method, which
is implemented in RDKit.
If it turns out to be efficient, I'd like to contribute to RDKit in
addition to Open Babel.
I heard some people are working on the distance geometry code.
If you give me any suggestions, I'd appreciate it.

You can read my proposal for more details:
https://drive.google.com/file/d/14xUVci1dbcrwCeWc088votLBZxp5Jiq8/

Project abstract:

Open Babel is a widely used open source toolkit for cheminformatics.
One of its important functions is 3D structure prediction of input
molecules. Open Babel’s coordinate generation function was improved in
speed and stereochemical accuracy by the new fragment-based coordinate
generation method implemented as a GSoC 2018 project. However, the
stereochemistry accuracy is not as high as the distance geometry
method of RDKit.

In this project, I will improve the coordinate generation of Open
Babel by implementing a new method which combines the fragment-based
method of Open Babel and the distance geometry method of RDKit. This
new implementation is expected to be faster than current RDKit and
more accurate in stereochemistry than current Open Babel. Implementing
a better prediction method could be beneficial for a wide range of
applications such as drug design.

Thanks,
Naruki

---
Naruki Yoshikawa
Department of Computational Biology and Medical Sciences,
Graduate School of Frontier Sciences
The University of Tokyo

Re: [Rdkit-devel] [Rdkit-announce] 2019.03.1 RDKit Release

[Greg L. <gre...@gm...>]

Thanks for that feedback; when I am back in the office this weekend I will
take a look at the Mac conda  install and see if we can fix this with some
specified dependencies

On Wed, 10 Apr 2019 at 01:20, Drew Gibson <dre...@go...>
wrote:

> Hi,
>
> I'm using the 2019-03.1 conda build of the new version of rdkit-postgresql
> on macOS Mohave 10.14.4.
>
> just FYI, I got the following error when trying to create the cartridge
> extension in a postgresql chembl_25 instance in the conda build -
>
> chembl_25=# create extension if not exists rdkit;
> ERROR:  could not load library
> "/Users/drew/anaconda3/envs/rdkit-psql-2019/lib/rdkit.so":
> dlopen(/Users/drew/anaconda3/envs/rdkit-psql-2019/lib/rdkit.so, 10):
> Library not loaded: @rpath/libboost_system.dylib
>   Referenced from: /Users/drew/anaconda3/envs/rdkit-psql-2019/lib/rdkit.so
>   Reason: image not found
>
> I couldn't see "libboost_system.dylib" in my conda environment /lib
> directory, so I installed anaconda boost and it fixed the problem - (from
> within the conda environment)
>
> conda install -c anaconda boost
>
> chembl_25=# create extension if not exists rdkit;
> CREATE EXTENSION
> chembl_25=#
>
> I then followed the other commands in the rdkit docs to create some
> fingerprints and build the similarity search index - all seems to be
> working ok.
>
> Cheers,  Drew
>
>
>
>
>
> On Tue, 9 Apr 2019 at 04:42, Greg Landrum <gre...@gm...> wrote:
>
>> Dear all,
>>
>> I'm pleased to announce that the next version of the RDKit - 2019.03 - is
>> released. The release notes are below.
>>
>> The release files are on the github release page:
>> https://github.com/rdkit/rdkit/releases/tag/Release_2019_03_1
>>
>> Binaries have been uploaded to anaconda.org (https://anaconda.org/rdkit).
>> The available conda binaries for this release are:
>> Linux 64bit: python 3.6, 3.7
>> Mac OS 64bit: python 3.6, 3.7
>> Windows 64bit: python 3.6, 3.7
>>
>> I believe that conda-forge will also switch to the new version in the
>> near future.
>>
>> Please note that the RDKit no longer supports Python 2.7. More details on
>> this here:
>>
>> https://www.mail-archive.com/rdk...@li.../msg08354.html
>>
>> I plan to put conda builds of the PostgreSQL cartridge up in the near
>> future.
>>
>> The online version of the documentation at rdkit.org (
>> http://rdkit.org/docs/index.html) has been updated.
>>
>> Some things that will be finished over the next couple of days:
>> - The conda build scripts will be updated to reflect the new version
>> - The homebrew script
>>
>> Thanks to everyone who submitted code, bug reports, and suggestions for
>> this release!
>>
>> Please let me know if you find any problems with the release or have
>> suggestions for the next one, which is scheduled for October 2019.
>>
>> Best Regards,
>> -greg
>>
>> # Release_2019.03.1
>> (Changes relative to Release_2018.09.1)
>>
>> ## REALLY IMPORTANT ANNOUNCEMENT
>> - As of this realease (2019.03.1) the RDKit no longer supports Python 2.
>> Please
>>   read this rdkit-discuss post to learn what your options are if you need
>> to
>>   keep using Python 2:
>>
>> https://www.mail-archive.com/rdk...@li.../msg08354.html
>>
>> ## Backwards incompatible changes
>> - The fix for github #2245 means that the default behavior of the
>> MaxMinPicker
>>   is now truly random. If you would like to reproduce the previous
>> behavior,
>>   provide a seed value of 42.
>> - The uncharging method in the MolStandardizer now attempts to generate
>>   canonical results for a given molecule. This may result in different
>> output
>>   for some molecules.
>>
>> ## Highlights:
>> - There's now a Japanese translation of large parts of the RDKit
>> documentation
>> - SGroup data can now be read from and written to Mol/SDF files
>> - The enhanced stereo handling has been improved: the information is now
>>   accessible from Python, EnumerateStereoisomers takes advantage of it,
>> and it
>>   can be read from and written to CXSmiles
>>
>> ## Acknowledgements:
>> Michael Banck, Francois Berenger, Thomas Blaschke, Brian Cole, Andrew
>> Dalke,
>> Bakary N'tji Diallo, Guillaume Godin, Anne Hersey, Jan Holst Jensen,
>> Sunhwan Jo,
>> Brian Kelley, Petr Kubat, Karl Leswing, Susan Leung, John Mayfield, Adam
>> Moyer,
>> Dan Nealschneider, Noel O'Boyle, Stephen Roughley, Takayuki Serizawa,
>> Gianluca
>> Sforna, Ricardo Rodriguez Schmidt, Gianluca Sforna, Matt Swain, Paolo
>> Tosco,
>> Ricardo Vianello, 'John-Videogames', 'magattaca', 'msteijaert',
>> 'paconius',
>> 'sirbiscuit'
>>
>> ## Bug Fixes:
>>   - PgSQL: fix boolean definitions for Postgresql 11
>>  (github pull #2129 from pkubatrh)
>>   - update fingerprint tutorial notebook
>>  (github pull #2130 from greglandrum)
>>   - Fix typo in RecapHierarchyNode destructor
>>  (github pull #2137 from iwatobipen)
>>   - SMARTS roundtrip failure
>>  (github issue #2142 from mcs07)
>>   - Error thrown in rdMolStandardize.ChargeParent
>>  (github issue #2144 from paconius)
>>   - SMILES parsing inconsistency based on input order
>>  (github issue #2148 from coleb)
>>   - MolDraw2D: line width not in python wrapper
>>  (github issue #2149 from greglandrum)
>>   - Missing Python API Documentation
>>  (github issue #2158 from greglandrum)
>>   - PgSQL: mol_to_svg() changes input molecule.
>>  (github issue #2174 from janholstjensen)
>>   - Remove Unicode From AcidBasePair Name
>>  (github pull #2185 from lilleswing)
>>   - Inconsistent treatment of `[as]` in SMILES and SMARTS
>>  (github issue #2197 from greglandrum)
>>   - RGroupDecomposition fixes, keep userLabels more robust
>> onlyMatchAtRGroups
>>  (github pull #2202 from bp-kelley)
>>   - Fix TautomerTransform in operator=
>>  (github pull #2203 from bp-kelley)
>>   - testEnumeration hangs/takes where long on 32bit architectures
>>  (github issue #2209 from mbanck)
>>   - Silencing some Python 3 warning messages
>>  (github pull #2223 from coleb)
>>   - removeHs shouldn't remove atom lists
>>  (github issue #2224 from rvianello)
>>   - failure round-tripping mol block with Q atom
>>  (github issue #2225 from rvianello)
>>   - problem round-tripping mol files that include bond topology info
>>  (github issue #2229 from rvianello)
>>   - aromatic main-group atoms written to SMARTS incorrectly
>>  (github issue #2237 from greglandrum)
>>   - findPotentialStereoBonds() stopping too early
>>  (github issue #2244 from greglandrum)
>>   - MinMax Diversity picker seeding shows deterministic / non-random
>> behaviour
>>  (github issue #2245 from sroughley)
>>   - Fix to serialize binary strings
>>  (github pull #2264 from bp-kelley)
>>   - Recognize N in three-membered rings as potentially chiral
>>  (github issue #2268 from greglandrum)
>>   - Failure when parsing mol block with M  PXA
>>  (github issue #2277 from greglandrum)
>>   - query-query matching failing for atoms constructed from SMARTS
>>  (github issue #2299 from greglandrum)
>>   - SMILES parsing fails for dative ring closures
>>  (github issue #2303 from greglandrum)
>>   - Missing Dict.h overload: std::string Dict::getVal<std::string>
>>  (github issue #2308 from greglandrum)
>>   - fix a problem with the random pickers test
>>  (github pull #2310 from greglandrum)
>>   - Some tests still failing on less common platforms.
>>  (github issue #2311 from giallu)
>>   - AddHs() using 3D coordinates with 2D conformations
>>  (github pull #2328 from greglandrum)
>>   - change to make the SWIG builds work on windows
>>  (github pull #2340 from greglandrum)
>>   - uncharger behaves differently on molecules constructed from mol
>> blocks and SMILES
>>  (github issue #2346 from greglandrum)
>>   - Memory Error When Writing ToBinary With "AllProps"
>>  (github issue #2352 from atom-moyer)
>>   - Seg fault on init if RDBASE is not set
>>  (github issue #2368 from greglandrum)
>>   - PandasTools.FrameToGridImage() fails with SVG output
>>  (github issue #2380 from greglandrum)
>>   - ClusterMols.GetDistanceMatrix throws a type error in Python 3
>>  (github issue #2387 from John-Videogames)
>>   - Uncharging logic reversed: protonate non-acids first
>>  (github issue #2392 from Anne Hersey)
>>
>> ## New Features and Enhancements:
>>   - Allow access to Enhanced Stereochemistry information from Python
>>  (github issue #2108 from d-b-w)
>>   - Adopt EnumerateStereoisomers to use extended stereo
>>  (github issue #2109 from greglandrum)
>>   - Enable ctest -T memcheck
>>  (github pull #2113 from ricrogz)
>>   - Support for parsing/writing SGroups in SD Mol files
>>  (github pull #2138 from ricrogz)
>>    - Rename the #define _DEBUG to MMPA_DEBUG in mmpa.cpp
>>  (github pull #2140 from baoilleach)
>>   - MolDraw2D: line width should be controlled by MolDrawOptions
>>  (github issue #2151 from greglandrum)
>>   - Some refactoring of the distance geometry code
>>  (github pull #2153 from greglandrum)
>>   - Less warnings
>>  (github pull #2155 from UnixJunkie)
>>   - ShapeTverskyIndex python function
>>  (github pull #2156 from susanhleung)
>>   - Skip compound if smiles conversion fails
>>  (github pull #2168 from msteijaert)
>>   - Fix #2176: InChI functions should return NULL on un-InChI-able input
>> molecules.
>>  (github pull #2177 from janholstjensen)
>>   - Update installation instructions for Linux
>>  (github pull #2181 from sirbiscuit)
>>   - Update CMake rules to find external coorgen & maeparser libs
>>  (github pull #2184 from ricrogz)
>>   - Update to use the travis Xenial environment
>>  (github pull #2200 from greglandrum)
>>   - Do not allow PandasTools to overwrite pandas settings
>>  (github pull #2206 from sirbiscuit)
>>   - re-enable (and update) the file parser tests
>>  (github pull #2208 from greglandrum)
>>   - Added documentation files written in Japanese into Book directory
>>  (github pull #2210 from magattaca)
>>   - Add C++ convenience function for drawing ROMols
>>  (github issue #2220 from greglandrum)
>>   - Change boost int types to std types
>>  (github pull #2233 from bp-kelley)
>>   - Added exports for SGroup functions
>>  (github pull #2242 from ricrogz)
>>   - Use coordMap when starting embedding from random coords
>>  (github issue #2246 from greglandrum)
>>   - Improve interactivity of output SVG
>>  (github pull #2253 from greglandrum)
>>   - Add options for substructure searching
>>  (github pull #2254 from greglandrum)
>>   - keep extra information about bonds from Mol files
>>  (github pull #2260 from greglandrum)
>>   - Allow converting mol blocks directly to InChI
>>  (github pull #2262 from greglandrum)
>>   - Patch/pains updates
>>  (github pull #2272 from johnmay)
>>   - add warning for 2D conformations flagged as 3D
>>  (github pull #2273 from greglandrum)
>>   - Store extra CXSMILES data as a property
>>  (github pull #2281 from ricrogz)
>>   - Parse enhanced stereo information from CXSMILES
>>  (github pull #2282 from ricrogz)
>>   - Robustify parsing of CTABs and SGROUPs
>>  (github pull #2283 from greglandrum)
>>   - Write enhanced stereo to cxsmiles
>>  (github pull #2290 from greglandrum)
>>   - Allow custom type-handlers in the RDProps interface
>>  (github pull #2293 from bp-kelley)
>>   - Add serialization to SubstructLibrary
>>  (github pull #2295 from bp-kelley)
>>   - support reading/writing atom props from SD files
>>  (github pull #2297 from greglandrum)
>>   - Add test for issue #2285, fix molbundle test
>>  (github pull #2301 from bp-kelley)
>>   - Update maeparser & coordgen libraries
>>  (github pull #2302 from ricrogz)
>>   - Mem errors clean up
>>  (github pull #2305 from ricrogz)
>>   - Add definition of MolFragmentToCXSmiles
>>  (github pull #2307 from greglandrum)
>>   - Doc update
>>  (github pull #2312 from greglandrum)
>>   - Adds gzstream stream, exposes to swig
>>  (github pull #2314 from bp-kelley)
>>   - Remove a bunch of Python2-related warts
>>  (github pull #2315 from greglandrum)
>>   - some much-needed optimization work on the new property lists
>>  (github pull #2317 from greglandrum)
>>   - Build warnings revisited
>>  (github pull #2318 from ricrogz)
>>   - change bogus "3D" to "2D" in a test file
>>  (github pull #2319 from greglandrum)
>>   - Allow copying atoms in Python
>>  (github pull #2322 from d-b-w)
>>   - fixes an r-group symmetrization problem
>>  (github pull #2324 from greglandrum)
>>   - simple docstring fix
>>  (github pull #2326 from sunhwan)
>>   - allow using system's catch2 for tests
>>  (github pull #2327 from giallu)
>>   - Python wrap DetectAtomStereoChemistry from MolFileStereochem.h
>>  (github issue #2329 from d-b-w)
>>   - switch to using cmake to handle the C++ spec
>>  (github pull #2334 from greglandrum)
>>   - WIP: optional integration with YAeHMOP
>>  (github pull #2335 from greglandrum)
>>   - Exposes substructlibrary to swig
>>  (github pull #2337 from bp-kelley)
>>   - Add a skip_all_if_match option to the FragmentRemover
>>  (github pull #2338 from greglandrum)
>>   - Dev/general csharp fixes
>>  (github pull #2341 from bp-kelley)
>>   - Add a read-only Python wrapper for SGroups
>>  (github pull #2343 from greglandrum)
>>   - Expose RGroupDecomposition to SWIG
>>  (github pull #2345 from greglandrum)
>>   - update debian build script to python3
>>  (github pull #2350 from UnixJunkie)
>>   - add GetStereoIsomerCount() function to EnumerateStereoisomers
>>  (github pull #2354 from greglandrum)
>>   - Update coordgenlibs to v1.2.2
>>  (github pull #2355 from ricrogz)
>>   - Small fixes to get DLLs to build on Windows
>>  (github pull #2356 from ptosco)
>>   - Boost deprecation warning
>>  (github pull #2357 from d-b-w)
>>   - Removes an extra debugging cerr statment
>>  (github pull #2360 from d-b-w)
>>   - Preserve enhanced stereo in reactions
>>  (github pull #2366 from d-b-w)
>>   - improvements to the Uncharge functionality
>>  (github pull #2374 from greglandrum)
>>   - Add ExplicitBitVect prop and query
>>  (github pull #2384 from bp-kelley)
>>   - Allow components of the MolStandardize code to be initialized from
>> streams
>>  (github pull #2385 from greglandrum)
>>
>> _______________________________________________
>> Rdkit-announce mailing list
>> Rdk...@li...
>> https://lists.sourceforge.net/lists/listinfo/rdkit-announce
>>
>

[Rdkit-devel] Canonical representation of Bond stereochemistry

[Dan W. <dan...@sc...>]

I'm curious about what I should be using to set and get double bond
stereochemistry. Is it the bond stereo of the double bond, or the bond
directions of the adjacent bonds?

I've been in-memory translating Mol objects into our internal
representation of a molecule. I've been using the Bond::getStereo() and
Bond::setStereo() to get and set bond stereochemistry.

Recently, I noticed that the products of RDKit reactions seem to
consistently not have bond stereo set. They have bond directions, though.

Here's a demonstration:

    from rdkit import Chem
    from rdkit.Chem import AllChem
    import rdkit.Chem.rdchem

    mol = Chem.MolFromSmiles(r'F/C=C/F')
    rxn = AllChem.ReactionFromSmarts('[C:1]>>[C:1]')
    product = rxn.RunReactants([mol])[0][0]
    assert mol.GetBondBetweenAtoms(1, 2).GetStereo() ==
rdkit.Chem.rdchem.BondStereo.STEREOE
    assert product.GetBondBetweenAtoms(0, 2).GetBondType() ==
rdkit.Chem.rdchem.BondType.DOUBLE

    print(Chem.MolToSmiles(product))
    print(product.GetBondBetweenAtoms(1, 0).GetBondDir(),
product.GetBondBetweenAtoms(0, 2).GetStereo(),
product.GetBondBetweenAtoms(2, 3).GetBondDir())
    # this fails:
    assert product.GetBondBetweenAtoms(0, 2).GetStereo() ==
rdkit.Chem.rdchem.BondStereo.STEREOE

I think that this means that I don't understand how I should be using bond
stereochemistry. It's also possible, though, that RxnRunner needs to assign
bond stereo from the bond directions after the reaction.

- dan nealschneider


Senior Developer
Schr*ö*dinger, Inc
Portland, OR

[Rdkit-devel] 2019.03.1 RDKit Release

[Greg L. <gre...@gm...>]

Dear all,

I'm pleased to announce that the next version of the RDKit - 2019.03 - is
released. The release notes are below.

The release files are on the github release page:
https://github.com/rdkit/rdkit/releases/tag/Release_2019_03_1

Binaries have been uploaded to anaconda.org (https://anaconda.org/rdkit).
The available conda binaries for this release are:
Linux 64bit: python 3.6, 3.7
Mac OS 64bit: python 3.6, 3.7
Windows 64bit: python 3.6, 3.7

I believe that conda-forge will also switch to the new version in the near
future.

Please note that the RDKit no longer supports Python 2.7. More details on
this here:
https://www.mail-archive.com/rdk...@li.../msg08354.html

I plan to put conda builds of the PostgreSQL cartridge up in the near
future.

The online version of the documentation at rdkit.org (
http://rdkit.org/docs/index.html) has been updated.

Some things that will be finished over the next couple of days:
- The conda build scripts will be updated to reflect the new version
- The homebrew script

Thanks to everyone who submitted code, bug reports, and suggestions for
this release!

Please let me know if you find any problems with the release or have
suggestions for the next one, which is scheduled for October 2019.

Best Regards,
-greg

# Release_2019.03.1
(Changes relative to Release_2018.09.1)

## REALLY IMPORTANT ANNOUNCEMENT
- As of this realease (2019.03.1) the RDKit no longer supports Python 2.
Please
  read this rdkit-discuss post to learn what your options are if you need to
  keep using Python 2:

https://www.mail-archive.com/rdk...@li.../msg08354.html

## Backwards incompatible changes
- The fix for github #2245 means that the default behavior of the
MaxMinPicker
  is now truly random. If you would like to reproduce the previous behavior,
  provide a seed value of 42.
- The uncharging method in the MolStandardizer now attempts to generate
  canonical results for a given molecule. This may result in different
output
  for some molecules.

## Highlights:
- There's now a Japanese translation of large parts of the RDKit
documentation
- SGroup data can now be read from and written to Mol/SDF files
- The enhanced stereo handling has been improved: the information is now
  accessible from Python, EnumerateStereoisomers takes advantage of it, and
it
  can be read from and written to CXSmiles

## Acknowledgements:
Michael Banck, Francois Berenger, Thomas Blaschke, Brian Cole, Andrew Dalke,
Bakary N'tji Diallo, Guillaume Godin, Anne Hersey, Jan Holst Jensen,
Sunhwan Jo,
Brian Kelley, Petr Kubat, Karl Leswing, Susan Leung, John Mayfield, Adam
Moyer,
Dan Nealschneider, Noel O'Boyle, Stephen Roughley, Takayuki Serizawa,
Gianluca
Sforna, Ricardo Rodriguez Schmidt, Gianluca Sforna, Matt Swain, Paolo
Tosco,
Ricardo Vianello, 'John-Videogames', 'magattaca', 'msteijaert', 'paconius',
'sirbiscuit'

## Bug Fixes:
  - PgSQL: fix boolean definitions for Postgresql 11
 (github pull #2129 from pkubatrh)
  - update fingerprint tutorial notebook
 (github pull #2130 from greglandrum)
  - Fix typo in RecapHierarchyNode destructor
 (github pull #2137 from iwatobipen)
  - SMARTS roundtrip failure
 (github issue #2142 from mcs07)
  - Error thrown in rdMolStandardize.ChargeParent
 (github issue #2144 from paconius)
  - SMILES parsing inconsistency based on input order
 (github issue #2148 from coleb)
  - MolDraw2D: line width not in python wrapper
 (github issue #2149 from greglandrum)
  - Missing Python API Documentation
 (github issue #2158 from greglandrum)
  - PgSQL: mol_to_svg() changes input molecule.
 (github issue #2174 from janholstjensen)
  - Remove Unicode From AcidBasePair Name
 (github pull #2185 from lilleswing)
  - Inconsistent treatment of `[as]` in SMILES and SMARTS
 (github issue #2197 from greglandrum)
  - RGroupDecomposition fixes, keep userLabels more robust
onlyMatchAtRGroups
 (github pull #2202 from bp-kelley)
  - Fix TautomerTransform in operator=
 (github pull #2203 from bp-kelley)
  - testEnumeration hangs/takes where long on 32bit architectures
 (github issue #2209 from mbanck)
  - Silencing some Python 3 warning messages
 (github pull #2223 from coleb)
  - removeHs shouldn't remove atom lists
 (github issue #2224 from rvianello)
  - failure round-tripping mol block with Q atom
 (github issue #2225 from rvianello)
  - problem round-tripping mol files that include bond topology info
 (github issue #2229 from rvianello)
  - aromatic main-group atoms written to SMARTS incorrectly
 (github issue #2237 from greglandrum)
  - findPotentialStereoBonds() stopping too early
 (github issue #2244 from greglandrum)
  - MinMax Diversity picker seeding shows deterministic / non-random
behaviour
 (github issue #2245 from sroughley)
  - Fix to serialize binary strings
 (github pull #2264 from bp-kelley)
  - Recognize N in three-membered rings as potentially chiral
 (github issue #2268 from greglandrum)
  - Failure when parsing mol block with M  PXA
 (github issue #2277 from greglandrum)
  - query-query matching failing for atoms constructed from SMARTS
 (github issue #2299 from greglandrum)
  - SMILES parsing fails for dative ring closures
 (github issue #2303 from greglandrum)
  - Missing Dict.h overload: std::string Dict::getVal<std::string>
 (github issue #2308 from greglandrum)
  - fix a problem with the random pickers test
 (github pull #2310 from greglandrum)
  - Some tests still failing on less common platforms.
 (github issue #2311 from giallu)
  - AddHs() using 3D coordinates with 2D conformations
 (github pull #2328 from greglandrum)
  - change to make the SWIG builds work on windows
 (github pull #2340 from greglandrum)
  - uncharger behaves differently on molecules constructed from mol blocks
and SMILES
 (github issue #2346 from greglandrum)
  - Memory Error When Writing ToBinary With "AllProps"
 (github issue #2352 from atom-moyer)
  - Seg fault on init if RDBASE is not set
 (github issue #2368 from greglandrum)
  - PandasTools.FrameToGridImage() fails with SVG output
 (github issue #2380 from greglandrum)
  - ClusterMols.GetDistanceMatrix throws a type error in Python 3
 (github issue #2387 from John-Videogames)
  - Uncharging logic reversed: protonate non-acids first
 (github issue #2392 from Anne Hersey)

## New Features and Enhancements:
  - Allow access to Enhanced Stereochemistry information from Python
 (github issue #2108 from d-b-w)
  - Adopt EnumerateStereoisomers to use extended stereo
 (github issue #2109 from greglandrum)
  - Enable ctest -T memcheck
 (github pull #2113 from ricrogz)
  - Support for parsing/writing SGroups in SD Mol files
 (github pull #2138 from ricrogz)
   - Rename the #define _DEBUG to MMPA_DEBUG in mmpa.cpp
 (github pull #2140 from baoilleach)
  - MolDraw2D: line width should be controlled by MolDrawOptions
 (github issue #2151 from greglandrum)
  - Some refactoring of the distance geometry code
 (github pull #2153 from greglandrum)
  - Less warnings
 (github pull #2155 from UnixJunkie)
  - ShapeTverskyIndex python function
 (github pull #2156 from susanhleung)
  - Skip compound if smiles conversion fails
 (github pull #2168 from msteijaert)
  - Fix #2176: InChI functions should return NULL on un-InChI-able input
molecules.
 (github pull #2177 from janholstjensen)
  - Update installation instructions for Linux
 (github pull #2181 from sirbiscuit)
  - Update CMake rules to find external coorgen & maeparser libs
 (github pull #2184 from ricrogz)
  - Update to use the travis Xenial environment
 (github pull #2200 from greglandrum)
  - Do not allow PandasTools to overwrite pandas settings
 (github pull #2206 from sirbiscuit)
  - re-enable (and update) the file parser tests
 (github pull #2208 from greglandrum)
  - Added documentation files written in Japanese into Book directory
 (github pull #2210 from magattaca)
  - Add C++ convenience function for drawing ROMols
 (github issue #2220 from greglandrum)
  - Change boost int types to std types
 (github pull #2233 from bp-kelley)
  - Added exports for SGroup functions
 (github pull #2242 from ricrogz)
  - Use coordMap when starting embedding from random coords
 (github issue #2246 from greglandrum)
  - Improve interactivity of output SVG
 (github pull #2253 from greglandrum)
  - Add options for substructure searching
 (github pull #2254 from greglandrum)
  - keep extra information about bonds from Mol files
 (github pull #2260 from greglandrum)
  - Allow converting mol blocks directly to InChI
 (github pull #2262 from greglandrum)
  - Patch/pains updates
 (github pull #2272 from johnmay)
  - add warning for 2D conformations flagged as 3D
 (github pull #2273 from greglandrum)
  - Store extra CXSMILES data as a property
 (github pull #2281 from ricrogz)
  - Parse enhanced stereo information from CXSMILES
 (github pull #2282 from ricrogz)
  - Robustify parsing of CTABs and SGROUPs
 (github pull #2283 from greglandrum)
  - Write enhanced stereo to cxsmiles
 (github pull #2290 from greglandrum)
  - Allow custom type-handlers in the RDProps interface
 (github pull #2293 from bp-kelley)
  - Add serialization to SubstructLibrary
 (github pull #2295 from bp-kelley)
  - support reading/writing atom props from SD files
 (github pull #2297 from greglandrum)
  - Add test for issue #2285, fix molbundle test
 (github pull #2301 from bp-kelley)
  - Update maeparser & coordgen libraries
 (github pull #2302 from ricrogz)
  - Mem errors clean up
 (github pull #2305 from ricrogz)
  - Add definition of MolFragmentToCXSmiles
 (github pull #2307 from greglandrum)
  - Doc update
 (github pull #2312 from greglandrum)
  - Adds gzstream stream, exposes to swig
 (github pull #2314 from bp-kelley)
  - Remove a bunch of Python2-related warts
 (github pull #2315 from greglandrum)
  - some much-needed optimization work on the new property lists
 (github pull #2317 from greglandrum)
  - Build warnings revisited
 (github pull #2318 from ricrogz)
  - change bogus "3D" to "2D" in a test file
 (github pull #2319 from greglandrum)
  - Allow copying atoms in Python
 (github pull #2322 from d-b-w)
  - fixes an r-group symmetrization problem
 (github pull #2324 from greglandrum)
  - simple docstring fix
 (github pull #2326 from sunhwan)
  - allow using system's catch2 for tests
 (github pull #2327 from giallu)
  - Python wrap DetectAtomStereoChemistry from MolFileStereochem.h
 (github issue #2329 from d-b-w)
  - switch to using cmake to handle the C++ spec
 (github pull #2334 from greglandrum)
  - WIP: optional integration with YAeHMOP
 (github pull #2335 from greglandrum)
  - Exposes substructlibrary to swig
 (github pull #2337 from bp-kelley)
  - Add a skip_all_if_match option to the FragmentRemover
 (github pull #2338 from greglandrum)
  - Dev/general csharp fixes
 (github pull #2341 from bp-kelley)
  - Add a read-only Python wrapper for SGroups
 (github pull #2343 from greglandrum)
  - Expose RGroupDecomposition to SWIG
 (github pull #2345 from greglandrum)
  - update debian build script to python3
 (github pull #2350 from UnixJunkie)
  - add GetStereoIsomerCount() function to EnumerateStereoisomers
 (github pull #2354 from greglandrum)
  - Update coordgenlibs to v1.2.2
 (github pull #2355 from ricrogz)
  - Small fixes to get DLLs to build on Windows
 (github pull #2356 from ptosco)
  - Boost deprecation warning
 (github pull #2357 from d-b-w)
  - Removes an extra debugging cerr statment
 (github pull #2360 from d-b-w)
  - Preserve enhanced stereo in reactions
 (github pull #2366 from d-b-w)
  - improvements to the Uncharge functionality
 (github pull #2374 from greglandrum)
  - Add ExplicitBitVect prop and query
 (github pull #2384 from bp-kelley)
  - Allow components of the MolStandardize code to be initialized from
streams
 (github pull #2385 from greglandrum)

Re: [Rdkit-devel] 2019_03_1b1: lots of undefined references to libboost_iostreams functions

[<bal...@un...>]

hi

thanks for taking the time to reply
Meanwhile I have investigated a bit and hopefully clarified to some extent
the situation (see my reply to greg)

> I believe the problem here is that you built boost without zlib support. 
> In the b2 boost build step, check that zlib is set to "yes" (it will be 

I have checked your suggestion and it looks to me that zlib detection by
boost is ok:

    ----8<----
    notice: [zlib] Using pre-installed library
    notice: [zlib] Condition
    [...]
        - zlib                     : yes
    ---->8----

> if the zlib devel package was installed on your machine), otherwise 
> things won't work, as boost_iostreams will be built without zlib support.

In the end, it looks to me that the problem is with the cmake related
changes in boost-1_70_0_rc1 vs previous boost versions.

thanks
ciao
-gabriele

Re: [Rdkit-devel] 2019_03_1b1: lots of undefined references to libboost_iostreams functions

[<bal...@un...>]

hi

> I haven't tried a boost 1.70 build yet, so I can't confirm that it works.
> Could you please try 1.69 and see if that helps?
> If not, please send the exact cmake command you used and what the output
> was.

thanks a lot for the quick reply.

Apologies: for some reason I was convinced that it had to be some
simple typo or similar problem and didn't enclose any detail

And: yes, building vs boost-1.69.0 works smoothly as always. So
problem seems to be building vs 1_70_0_rc1 (which is a run candidate,
after all!).

For completeness I enclose below my cmake call and corresponding output vs
both 1.69.0 and 1_70_0_rc1.

BTW, I've also investigated in some more detail the files installed by
boost and found that 1_70_0_rc1 installs a lot of .cmake files where
1.69.0 and previous didn't install any (at least here on my machine)

Also, there are some new .cmake files in the 1_70_0_rc1 tarball that
were not present in 1.69.0 (list enclosed below)

In the end, it seems to me that there are no problems with rdkit when
built vs a *stable* version of boost: apologies for the noise...


ciao
-gabriele


cmake output when boost-1.69.0
==============================
cmake .. -G Ninja -DCMAKE_INSTALL_PREFIX=/opt/stow.d/versions/rdkit-2019_03_1b1/usr -DRDK_INSTALL_INTREE=OFF -DLIB_SUFFIX=64 -DPYTHON_INSTDIR=/opt/stow.d/versions/rdkit-2019_03_1b1/usr/python3/site_python3/lib64 -DBoost_INCLUDE_DIR:PATH=/usr/include -DBoost_LIBRARY_DIR_DEBUG:PATH=/usr/lib64 -DBoost_LIBRARY_DIR_RELEASE:PATH=/usr/lib64 -DPYTHON_EXECUTABLE:FILEPATH=/usr/python3/bin/python -DRDK_BOOST_PYTHON3_NAME:STRING=python38
-- Found Catch2 source in /home/balducci/tmp/install-us-d/rdkit-2019_03_1b1.d/rdkit-2019_03_1b1/External/catch/catch
CATCH: /home/balducci/tmp/install-us-d/rdkit-2019_03_1b1.d/rdkit-2019_03_1b1/External/catch/catch/single_include
Python Install directory /opt/stow.d/versions/rdkit-2019_03_1b1/usr/python3/site_python3/lib64
PYTHON Py_ENABLE_SHARED: 1
PYTHON USING LINK LINE: -m64 -shared -pthread -L/usr/Xorg/lib64 -pthread -L/usr/Xorg/lib64
-- Boost version: 1.69.0
-- Found the following Boost libraries:
--   serialization
== Using strict rotor definition
-- maeparser include dir set as '/home/balducci/tmp/install-us-d/rdkit-2019_03_1b1.d/rdkit-2019_03_1b1/External/CoordGen'
-- maeparser libraries set as 'maeparser_LIBRARIES-NOTFOUND'
-- Could NOT find maeparser (missing: maeparser_LIBRARIES) 
-- Found MAEParser source in /home/balducci/tmp/install-us-d/rdkit-2019_03_1b1.d/rdkit-2019_03_1b1/External/CoordGen/maeparser
-- Boost version: 1.69.0
-- Found the following Boost libraries:
--   system
--   iostreams
--   regex
-- coordgen include dir set as /home/balducci/tmp/install-us-d/rdkit-2019_03_1b1.d/rdkit-2019_03_1b1/External/CoordGen
-- coordgen libraries set as 'coordgen_LIBRARIES-NOTFOUND'
-- coordgen templates file set as 'coordgen_TEMPLATE_FILE-NOTFOUND'
-- Could NOT find coordgen (missing: coordgen_LIBRARIES coordgen_TEMPLATE_FILE) 
-- Found coordgenlibs source in /home/balducci/tmp/install-us-d/rdkit-2019_03_1b1.d/rdkit-2019_03_1b1/External/CoordGen/coordgen
-- Boost version: 1.69.0
-- Found the following Boost libraries:
--   system
--   iostreams
--   regex
-- Boost version: 1.69.0
-- Found the following Boost libraries:
--   system
--   iostreams
--   regex
== Updating Filters.cpp from pains file
== Done updating pains files
-- Found RapidJSON source in /home/balducci/tmp/install-us-d/rdkit-2019_03_1b1.d/rdkit-2019_03_1b1/External
-- Configuring done
-- Generating done
-- Build files have been written to: /home/balducci/tmp/install-us-d/rdkit-2019_03_1b1.d/rdkit-2019_03_1b1/build


cmake output when boost-1_70_0_rc1
==================================
cmake .. -G Ninja -DCMAKE_INSTALL_PREFIX=/opt/stow.d/versions/rdkit-2019_03_1b1/usr -DRDK_INSTALL_INTREE=OFF -DLIB_SUFFIX=64 -DPYTHON_INSTDIR=/opt/stow.d/versions/rdkit-2019_03_1b1/usr/python3/site_python3/lib64 -DBoost_INCLUDE_DIR:PATH=/usr/include -DBoost_LIBRARY_DIR_DEBUG:PATH=/usr/lib64 -DBoost_LIBRARY_DIR_RELEASE:PATH=/usr/lib64 -DPYTHON_EXECUTABLE:FILEPATH=/usr/python3/bin/python -DRDK_BOOST_PYTHON3_NAME:STRING=python38
-- Found Catch2 source in /home/balducci/tmp/install-us-d/rdkit-2018_09_3.d/rdkit-2019_03_1b1/External/catch/catch
CATCH: /home/balducci/tmp/install-us-d/rdkit-2018_09_3.d/rdkit-2019_03_1b1/External/catch/catch/single_include
-- Found Boost 1.70.0 at /usr/lib64/cmake/Boost-1.70.0
--   Requested configuration: QUIET COMPONENTS python38
-- Found boost_headers 1.70.0 at /usr/lib64/cmake/boost_headers-1.70.0
-- Boost 1.56.0 found.
-- Found Boost components:
   python38
Python Install directory /opt/stow.d/versions/rdkit-2019_03_1b1/usr/python3/site_python3/lib64
PYTHON Py_ENABLE_SHARED: 1
PYTHON USING LINK LINE: -m64 -shared -pthread -L/usr/Xorg/lib64 -pthread -L/usr/Xorg/lib64
-- Found Boost 1.70.0 at /usr/lib64/cmake/Boost-1.70.0
--   Requested configuration: QUIET COMPONENTS serialization
-- Found boost_serialization 1.70.0 at /usr/lib64/cmake/boost_serialization-1.70.0
--   libboost_serialization.a
-- Adding boost_serialization dependencies: headers
-- Boost 1.56.0 found.
-- Found Boost components:
   serialization
== Using strict rotor definition
-- maeparser include dir set as '/home/balducci/tmp/install-us-d/rdkit-2018_09_3.d/rdkit-2019_03_1b1/External/CoordGen'
-- maeparser libraries set as 'maeparser_LIBRARIES-NOTFOUND'
-- Could NOT find maeparser (missing: maeparser_LIBRARIES) 
-- Found MAEParser source in /home/balducci/tmp/install-us-d/rdkit-2018_09_3.d/rdkit-2019_03_1b1/External/CoordGen/maeparser
-- Found Boost 1.70.0 at /usr/lib64/cmake/Boost-1.70.0
--   Requested configuration: QUIET REQUIRED COMPONENTS system;iostreams
-- Found boost_system 1.70.0 at /usr/lib64/cmake/boost_system-1.70.0
--   libboost_system.a
-- Adding boost_system dependencies: headers
-- Found boost_iostreams 1.70.0 at /usr/lib64/cmake/boost_iostreams-1.70.0
--   libboost_iostreams.a
-- Adding boost_iostreams dependencies: headers
-- Boost 1.56.0 found.
-- Found Boost components:
   system;iostreams
-- coordgen include dir set as /home/balducci/tmp/install-us-d/rdkit-2018_09_3.d/rdkit-2019_03_1b1/External/CoordGen
-- coordgen libraries set as 'coordgen_LIBRARIES-NOTFOUND'
-- coordgen templates file set as 'coordgen_TEMPLATE_FILE-NOTFOUND'
-- Could NOT find coordgen (missing: coordgen_LIBRARIES coordgen_TEMPLATE_FILE) 
-- Found coordgenlibs source in /home/balducci/tmp/install-us-d/rdkit-2018_09_3.d/rdkit-2019_03_1b1/External/CoordGen/coordgen
-- Found Boost 1.70.0 at /usr/lib64/cmake/Boost-1.70.0
--   Requested configuration: QUIET REQUIRED COMPONENTS system;iostreams
-- Found boost_system 1.70.0 at /usr/lib64/cmake/boost_system-1.70.0
--   libboost_system.a
-- Adding boost_system dependencies: headers
-- Found boost_iostreams 1.70.0 at /usr/lib64/cmake/boost_iostreams-1.70.0
--   libboost_iostreams.a
-- Adding boost_iostreams dependencies: headers
-- Boost 1.56.0 found.
-- Found Boost components:
   system;iostreams
-- Found Boost 1.70.0 at /usr/lib64/cmake/Boost-1.70.0
--   Requested configuration: QUIET REQUIRED COMPONENTS system;iostreams
-- Found boost_system 1.70.0 at /usr/lib64/cmake/boost_system-1.70.0
--   libboost_system.a
-- Adding boost_system dependencies: headers
-- Found boost_iostreams 1.70.0 at /usr/lib64/cmake/boost_iostreams-1.70.0
--   libboost_iostreams.a
-- Adding boost_iostreams dependencies: headers
-- Boost 1.56.0 found.
-- Found Boost components:
   system;iostreams
== Making EnumerateLibrary without boost Serialization support
== Making FilterCatalog without boost Serialization support
== Updating Filters.cpp from pains file
== Done updating pains files
== Making SubstructLibrary without boost Serialization support
-- Found RapidJSON source in /home/balducci/tmp/install-us-d/rdkit-2018_09_3.d/rdkit-2019_03_1b1/External
-- Configuring done
-- Generating done
-- Build files have been written to: /home/balducci/tmp/install-us-d/rdkit-2018_09_3.d/rdkit-2019_03_1b1/build


              .cmake files in  boost_1_69_0 and boost-1_70_0_rc1
              --------------------------------------------------
boost_1_69_0                                                    boost-1_70_0_rc1
=============================================================================================================================
							      >	./tools/boost_install/BoostDetectToolset.cmake
							      >	./tools/boost_install/test/BoostVersion.cmake
							      >	./tools/boost_install/test/iostreams/zlib-1.2.11/zlib.pc.cmak
							      >	./tools/boost_install/test/iostreams/zlib-1.2.11/zconf.h.cmak
							      >	./tools/boost_install/BoostConfig.cmake
							      >	./libs/smart_ptr/test/cmake_subdir_test
./libs/gil/example/cmake					./libs/gil/example/cmake
./libs/mp11/test/check_cmake_version.cpp			./libs/mp11/test/check_cmake_version.cpp
./libs/mp11/test/cmake_subdir_test				./libs/mp11/test/cmake_subdir_test
./libs/mp11/test/cmake_install_test				./libs/mp11/test/cmake_install_test
./libs/hof/cmake						./libs/hof/cmake
./libs/hof/cmake/SphinxDoc.cmake				./libs/hof/cmake/SphinxDoc.cmake
./libs/hof/cmake/BCMTest.cmake					./libs/hof/cmake/BCMTest.cmake
./libs/vmd/test/test_cmake					./libs/vmd/test/test_cmake
./libs/compute/cmake						./libs/compute/cmake
./libs/compute/cmake/FindEigen.cmake				./libs/compute/cmake/FindEigen.cmake
./libs/compute/cmake/FindBolt.cmake				./libs/compute/cmake/FindBolt.cmake
./libs/compute/cmake/BoostComputeConfig.cmake.in		./libs/compute/cmake/BoostComputeConfig.cmake.in
./libs/compute/cmake/opencl/FindOpenCL.cmake			./libs/compute/cmake/opencl/FindOpenCL.cmake
./libs/compute/cmake/FindTBB.cmake				./libs/compute/cmake/FindTBB.cmake
							      >	./libs/endian/test/cmake_subdir_test
							      >	./libs/predef/test/test_cmake
							      >	./libs/atomic/test/test_cmake
./libs/hana/example/cmake_integration				./libs/hana/example/cmake_integration
./libs/hana/cmake						./libs/hana/cmake
./libs/hana/cmake/TestHeaders.cmake				./libs/hana/cmake/TestHeaders.cmake
./libs/hana/cmake/FindMeta.cmake				./libs/hana/cmake/FindMeta.cmake
./libs/hana/cmake/CheckCxxCompilerSupport.cmake			./libs/hana/cmake/CheckCxxCompilerSupport.cmake
./libs/hana/cmake/FindMPL11.cmake				./libs/hana/cmake/FindMPL11.cmake
							      >	./libs/system/test/cmake_subdir_test
./libs/assert/test/check_cmake_version.cpp			./libs/assert/test/check_cmake_version.cpp
							      >	./libs/function/test/cmake_subdir_test
./libs/yap/cmake						./libs/yap/cmake
./libs/yap/cmake/Disassemble.cmake				./libs/yap/cmake/Disassemble.cmake
./libs/yap/cmake/dependencies.cmake				./libs/yap/cmake/dependencies.cmake
./libs/beast/test/extern/zlib-1.2.11/zlib.pc.cmakein		./libs/beast/test/extern/zlib-1.2.11/zlib.pc.cmakein
./libs/beast/test/extern/zlib-1.2.11/zconf.h.cmakein		./libs/beast/test/extern/zlib-1.2.11/zconf.h.cmakein
							      >	./libs/bind/test/cmake_subdir_test
							      >	./libs/core/test/cmake_subdir_test
							      >	./libs/array/test/test_cmake

Re: [Rdkit-devel] 2019_03_1b1: lots of undefined references to libboost_iostreams functions

[Paolo T. <pao...@gm...>]

Hi Gabriele,

I believe the problem here is that you built boost without zlib support. 
In the b2 boost build step, check that zlib is set to "yes" (it will be 
if the zlib devel package was installed on your machine), otherwise 
things won't work, as boost_iostreams will be built without zlib support.

Cheers,
p.

On 06/04/2019 14:52, bal...@un... wrote:
> hello,
>
> there seems to be a glitch in the 2019_03_1b1 build process, where a lot
> of link steps fail for me with undefined references to
> libboost_iostreams functions. Eg:
>
>      ----8<----
>      /usr/bin/ld: lib/libRDKitmaeparser.so.1.2019.03.1b1: undefined reference to `boost::iostreams::zlib::deflated'
>      /usr/bin/ld: lib/libRDKitmaeparser.so.1.2019.03.1b1: undefined reference to `boost::iostreams::detail::zlib_base::before(char const*&, char const*, char*&, char*)'
>
>      [...]
>
>      /usr/bin/ld: lib/libRDKitmaeparser.so.1.2019.03.1b1: undefined reference to `boost::iostreams::zlib::default_compression'
>      collect2: error: ld returned 1 exit status
>      ---->8----
>
> I can make my build complete successfully only with:
>      export LDFLAGS="-lboost_iostreams"
> Is this to be expected or is there a problem with the build process?
> (all previous releases didn't need that)
>
> I can add that ./External/CoordGen/CMakeLists.txt
> ./Code/GraphMol/FileParsers/CMakeLists.txt and
> ./Code/RDStreams/CMakeLists.txt do actually contain a:
>     set (link_iostreams ${Boost_LIBRARIES})
> definition, but apparently link_iostreams is eventually empty (but I
> might not have followed correctly the logic of the cmake files)
>
> Building on linux vs boost-1_70_0_rc1
>
> thanks a lot
> ciao
> gabriele
>
>
> _______________________________________________
> Rdkit-devel mailing list
> Rdk...@li...
> https://lists.sourceforge.net/lists/listinfo/rdkit-devel

Re: [Rdkit-devel] 2019_03_1b1: lots of undefined references to libboost_iostreams functions

[Greg L. <gre...@gm...>]

Hi Gabriele,

I haven't tried a boost 1.70 build yet, so I can't confirm that it works.
Could you please try 1.69 and see if that helps?
If not, please send the exact cmake command you used and what the output
was.

Thanks,
-greg


On Sat, Apr 6, 2019 at 3:53 PM <bal...@un...> wrote:

> hello,
>
> there seems to be a glitch in the 2019_03_1b1 build process, where a lot
> of link steps fail for me with undefined references to
> libboost_iostreams functions. Eg:
>
>     ----8<----
>     /usr/bin/ld: lib/libRDKitmaeparser.so.1.2019.03.1b1: undefined
> reference to `boost::iostreams::zlib::deflated'
>     /usr/bin/ld: lib/libRDKitmaeparser.so.1.2019.03.1b1: undefined
> reference to `boost::iostreams::detail::zlib_base::before(char const*&,
> char const*, char*&, char*)'
>
>     [...]
>
>     /usr/bin/ld: lib/libRDKitmaeparser.so.1.2019.03.1b1: undefined
> reference to `boost::iostreams::zlib::default_compression'
>     collect2: error: ld returned 1 exit status
>     ---->8----
>
> I can make my build complete successfully only with:
>     export LDFLAGS="-lboost_iostreams"
> Is this to be expected or is there a problem with the build process?
> (all previous releases didn't need that)
>
> I can add that ./External/CoordGen/CMakeLists.txt
> ./Code/GraphMol/FileParsers/CMakeLists.txt and
> ./Code/RDStreams/CMakeLists.txt do actually contain a:
>    set (link_iostreams ${Boost_LIBRARIES})
> definition, but apparently link_iostreams is eventually empty (but I
> might not have followed correctly the logic of the cmake files)
>
> Building on linux vs boost-1_70_0_rc1
>
> thanks a lot
> ciao
> gabriele
>
>
> _______________________________________________
> Rdkit-devel mailing list
> Rdk...@li...
> https://lists.sourceforge.net/lists/listinfo/rdkit-devel
>

[Rdkit-devel] 2019_03_1b1: lots of undefined references to libboost_iostreams functions

[<bal...@un...>]

hello,

there seems to be a glitch in the 2019_03_1b1 build process, where a lot
of link steps fail for me with undefined references to
libboost_iostreams functions. Eg:

    ----8<----
    /usr/bin/ld: lib/libRDKitmaeparser.so.1.2019.03.1b1: undefined reference to `boost::iostreams::zlib::deflated'
    /usr/bin/ld: lib/libRDKitmaeparser.so.1.2019.03.1b1: undefined reference to `boost::iostreams::detail::zlib_base::before(char const*&, char const*, char*&, char*)'

    [...]

    /usr/bin/ld: lib/libRDKitmaeparser.so.1.2019.03.1b1: undefined reference to `boost::iostreams::zlib::default_compression'
    collect2: error: ld returned 1 exit status
    ---->8----

I can make my build complete successfully only with:
    export LDFLAGS="-lboost_iostreams"
Is this to be expected or is there a problem with the build process?
(all previous releases didn't need that)

I can add that ./External/CoordGen/CMakeLists.txt
./Code/GraphMol/FileParsers/CMakeLists.txt and
./Code/RDStreams/CMakeLists.txt do actually contain a:
   set (link_iostreams ${Boost_LIBRARIES})
definition, but apparently link_iostreams is eventually empty (but I
might not have followed correctly the logic of the cmake files)

Building on linux vs boost-1_70_0_rc1

thanks a lot
ciao
gabriele

Re: [Rdkit-devel] [Rdkit-discuss] Beta of the 2019.03 release available

[Paolo T. <pao...@gm...>]

Dear Markus,

in case you ran into the Float8GetDatum undefined symbol issue (it just 
happened to me), please make sure that you are actually building the 
cartridge against the PostgreSQL 11 headers.

You may check this in the CMake output; for example, in my case it has 
to be:

postgres: /usr/pgsql-11/include;/usr/pgsql-11/include/server

and to obtain this I had to define in my cmake command line:

   -DPostgreSQL_ROOT=/usr/pgsql-11 \
   -DPostgreSQL_INCLUDE_DIR=/usr/pgsql-11/include \
-DPostgreSQL_TYPE_INCLUDE_DIR=/usr/pgsql-11/include/server \
   -DPostgreSQL_LIBRARY_DIR=/usr/pgsql-11/lib \

in addition to the usual

   -DRDK_BUILD_PGSQL=ON \
   -DRDK_PGSQL_STATIC=ON \

to avoid that FindPostgreSQL.cmake finds the headers of the system 
CentOS 7 PostgreSQL in /usr/include rather than the PostgreSQL headers 
in /usr/pgsql-11/include.

HTH, cheers,
p.

On 05/04/2019 17:42, Markus Sitzmann wrote:
> Hi Greg,
>
> my Chembience RDKit image build with version 2019.03-b1b went fine 
> (well, I just pull it with conda; in case someone is interested it is 
> available with tag 0.2.10-beta-1 at Dockerhub).
>
> For the Postgres extension (which I still compile myself during the 
> Docker build against Postgress), your python 3 enforcement uncovered 
> some dark corners of my build process, but that is fixed. However, 
> compiling 2019.03-b1b against Postgres 11 fails during compilation (am 
> I too cheeky?).
>
> Markus
>
> On Wed, Apr 3, 2019 at 11:38 AM Greg Landrum <gre...@gm... 
> <mailto:gre...@gm...>> wrote:
>
>     Dear all,
>
>     The beta of the 2019.03 RDKit release has been tagged in github:
>     https://github.com/rdkit/rdkit/releases/tag/Release_2019_03_1b1
>
>     There are a couple more bug fixes and maybe one more feature
>     expected before the actual release, but I wanted to go ahead and
>     get the beta out there.
>
>     I've done conda builds for Python 3.6 and 3.7 for Windows, Mac,
>     and Linux. These all use the beta label so that they do not
>     install by default; you'll need to run "conda install" as follows:
>
>     conda install -c rdkit/label/beta rdkit
>
>     Be sure to confirm that it's installing the right version when you
>     are prompted (if there's no build available, it will pick the
>     current production release instead).
>
>     The relevant section of the release notes is below, or you can see
>     a nicely formatted version here:
>     https://github.com/rdkit/rdkit/releases/tag/Release_2019_03_1b1
>
>     As usual, if you have time to try out the new release I would love
>     feedback. If nothing major comes up, I plan to do the actual
>     release early next week.
>
>     Best,
>     -greg
>
>     # Release_2019.03.1
>     (Changes relative to Release_2018.09.1)
>
>     ## REALLY IMPORTANT ANNOUNCEMENT
>     - As of this realease (2019.03.1) the RDKit no longer supports Python 2. Please read this rdkit-discuss post to learn what your options are if you need to keep using Python 2:
>        https://www.mail-archive.com/rdk...@li.../msg08354.html
>
>     ## Backwards incompatible changes
>     - The fix for github #2245 means that the default behavior of the MaxMinPicker
>        is now truly random. If you would like to reproduce the previous behavior,
>        provide a seed value of 42.
>     - The uncharging method in the MolStandardizer now attempts to generate
>        canonical results for a given molecule. This may result in different output
>        for some molecules.
>        
>     ## Highlights:
>     - There's now a Japanese translation of large parts of the RDKit documentation
>     - SGroup data can now be read from and written to Mol/SDF files
>     - The enhanced stereo handling has been improved: the information is now
>        accessible from Python, EnumerateStereoisomers takes advantage of it, and it
>        can be read from and written to CXSmiles
>
>     ## Acknowledgements:
>     Michael Banck, Francois Berenger, Thomas Blaschke, Brian Cole, Andrew Dalke,
>     Bakary N'tji Diallo, Guillaume Godin, Jan Holst Jensen, Sunhwan Jo, Brian
>     Kelley, Petr Kubat, Karl Leswing, Susan Leung, John Mayfield, Adam Moyer, Dan
>     Nealschneider, Noel O'Boyle, Stephen Roughley, Takayuki Serizawa, Gianluca
>     Sforna, Ricardo Rodriguez Schmidt, Matt Swain, Paolo Tosco, Ricardo Vianello,
>     'John-Videogames', 'magattaca', 'msteijaert', 'paconius', 'sirbiscuit'
>
>     ## Bug Fixes:
>        - PgSQL: fix boolean definitions for Postgresql 11
>       (github pull #2129 from pkubatrh)
>        - update fingerprint tutorial notebook
>       (github pull #2130 from greglandrum)
>        - Fix typo in RecapHierarchyNode destructor
>       (github pull #2137 from iwatobipen)
>        - SMARTS roundtrip failure
>       (github issue #2142 from mcs07)
>        - Error thrown in rdMolStandardize.ChargeParent
>       (github issue #2144 from paconius)
>        - SMILES parsing inconsistency based on input order
>       (github issue #2148 from coleb)
>        - MolDraw2D: line width not in python wrapper
>       (github issue #2149 from greglandrum)
>        - Missing Python API Documentation
>       (github issue #2158 from greglandrum)
>        - PgSQL: mol_to_svg() changes input molecule.
>       (github issue #2174 from janholstjensen)
>        - Remove Unicode From AcidBasePair Name
>       (github pull #2185 from lilleswing)
>        - Inconsistent treatment of `[as]` in SMILES and SMARTS
>       (github issue #2197 from greglandrum)
>        - RGroupDecomposition fixes, keep userLabels more robust onlyMatchAtRGroups
>       (github pull #2202 from bp-kelley)
>        - Fix TautomerTransform in operator=
>       (github pull #2203 from bp-kelley)
>        - testEnumeration hangs/takes where long on 32bit architectures
>       (github issue #2209 from mbanck)
>        - Silencing some Python 3 warning messages
>       (github pull #2223 from coleb)
>        - removeHs shouldn't remove atom lists
>       (github issue #2224 from rvianello)
>        - failure round-tripping mol block with Q atom
>       (github issue #2225 from rvianello)
>        - problem round-tripping mol files that include bond topology info
>       (github issue #2229 from rvianello)
>        - aromatic main-group atoms written to SMARTS incorrectly
>       (github issue #2237 from greglandrum)
>        - findPotentialStereoBonds() stopping too early
>       (github issue #2244 from greglandrum)
>        - MinMax Diversity picker seeding shows deterministic / non-random behaviour
>       (github issue #2245 from sroughley)
>        - Fix to serialize binary strings
>       (github pull #2264 from bp-kelley)
>        - Recognize N in three-membered rings as potentially chiral
>       (github issue #2268 from greglandrum)
>        - Failure when parsing mol block with M  PXA
>       (github issue #2277 from greglandrum)
>        - query-query matching failing for atoms constructed from SMARTS
>       (github issue #2299 from greglandrum)
>        - SMILES parsing fails for dative ring closures
>       (github issue #2303 from greglandrum)
>        - Missing Dict.h overload: std::string Dict::getVal<std::string>
>       (github issue #2308 from greglandrum)
>        - fix a problem with the random pickers test
>       (github pull #2310 from greglandrum)
>        - AddHs() using 3D coordinates with 2D conformations
>       (github pull #2328 from greglandrum)
>        - change to make the SWIG builds work on windows
>       (github pull #2340 from greglandrum)
>        - uncharger behaves differently on molecules constructed from mol blocks and SMILES
>       (github issue #2346 from greglandrum)
>        - Memory Error When Writing ToBinary With "AllProps"
>       (github issue #2352 from atom-moyer)
>        - Seg fault on init if RDBASE is not set
>       (github issue #2368 from greglandrum)
>        - PandasTools.FrameToGridImage() fails with SVG output
>       (github issue #2380 from greglandrum)
>        - ClusterMols.GetDistanceMatrix throws a type error in Python 3
>       (github issue #2387 from John-Videogames)
>
>     ## New Features and Enhancements:
>        - Allow access to Enhanced Stereochemistry information from Python
>       (github issue #2108 from d-b-w)
>        - Adopt EnumerateStereoisomers to use extended stereo
>       (github issue #2109 from greglandrum)
>        - Enable ctest -T memcheck
>       (github pull #2113 from ricrogz)
>        - Support for parsing/writing SGroups in SD Mol files
>       (github pull #2138 from ricrogz)
>         - Rename the #define _DEBUG to MMPA_DEBUG in mmpa.cpp
>       (github pull #2140 from baoilleach)
>        - MolDraw2D: line width should be controlled by MolDrawOptions
>       (github issue #2151 from greglandrum)
>        - Some refactoring of the distance geometry code
>       (github pull #2153 from greglandrum)
>        - Less warnings
>       (github pull #2155 from UnixJunkie)
>        - ShapeTverskyIndex python function
>       (github pull #2156 from susanhleung)
>        - Skip compound if smiles conversion fails
>       (github pull #2168 from msteijaert)
>        - Fix #2176: InChI functions should return NULL on un-InChI-able input molecules.
>       (github pull #2177 from janholstjensen)
>        - Update installation instructions for Linux
>       (github pull #2181 from sirbiscuit)
>        - Update CMake rules to find external coorgen & maeparser libs
>       (github pull #2184 from ricrogz)
>        - Update to use the travis Xenial environment
>       (github pull #2200 from greglandrum)
>        - Do not allow PandasTools to overwrite pandas settings
>       (github pull #2206 from sirbiscuit)
>        - re-enable (and update) the file parser tests
>       (github pull #2208 from greglandrum)
>        - Added documentation files written in Japanese into Book directory
>       (github pull #2210 from magattaca)
>        - Add C++ convenience function for drawing ROMols
>       (github issue #2220 from greglandrum)
>        - Change boost int types to std types
>       (github pull #2233 from bp-kelley)
>        - Added exports for SGroup functions
>       (github pull #2242 from ricrogz)
>        - Use coordMap when starting embedding from random coords
>       (github issue #2246 from greglandrum)
>        - Improve interactivity of output SVG
>       (github pull #2253 from greglandrum)
>        - Add options for substructure searching
>       (github pull #2254 from greglandrum)
>        - keep extra information about bonds from Mol files
>       (github pull #2260 from greglandrum)
>        - Allow converting mol blocks directly to InChI
>       (github pull #2262 from greglandrum)
>        - Patch/pains updates
>       (github pull #2272 from johnmay)
>        - add warning for 2D conformations flagged as 3D
>       (github pull #2273 from greglandrum)
>        - Store extra CXSMILES data as a property
>       (github pull #2281 from ricrogz)
>        - Parse enhanced stereo information from CXSMILES
>       (github pull #2282 from ricrogz)
>        - Robustify parsing of CTABs and SGROUPs
>       (github pull #2283 from greglandrum)
>        - Write enhanced stereo to cxsmiles
>       (github pull #2290 from greglandrum)
>        - Allow custom type-handlers in the RDProps interface
>       (github pull #2293 from bp-kelley)
>        - Add serialization to SubstructLibrary
>       (github pull #2295 from bp-kelley)
>        - support reading/writing atom props from SD files
>       (github pull #2297 from greglandrum)
>        - Add test for issue #2285, fix molbundle test
>       (github pull #2301 from bp-kelley)
>        - Update maeparser & coordgen libraries
>       (github pull #2302 from ricrogz)
>        - Mem errors clean up
>       (github pull #2305 from ricrogz)
>        - Add definition of MolFragmentToCXSmiles
>       (github pull #2307 from greglandrum)
>        - Doc update
>       (github pull #2312 from greglandrum)
>        - Adds gzstream stream, exposes to swig
>       (github pull #2314 from bp-kelley)
>        - Remove a bunch of Python2-related warts
>       (github pull #2315 from greglandrum)
>        - some much-needed optimization work on the new property lists
>       (github pull #2317 from greglandrum)
>        - Build warnings revisited
>       (github pull #2318 from ricrogz)
>        - change bogus "3D" to "2D" in a test file
>       (github pull #2319 from greglandrum)
>        - Allow copying atoms in Python
>       (github pull #2322 from d-b-w)
>        - fixes an r-group symmetrization problem
>       (github pull #2324 from greglandrum)
>        - simple docstring fix
>       (github pull #2326 from sunhwan)
>        - allow using system's catch2 for tests
>       (github pull #2327 from giallu)
>        - Python wrap DetectAtomStereoChemistry from MolFileStereochem.h
>       (github issue #2329 from d-b-w)
>        - switch to using cmake to handle the C++ spec
>       (github pull #2334 from greglandrum)
>        - WIP: optional integration with YAeHMOP
>       (github pull #2335 from greglandrum)
>        - Exposes substructlibrary to swig
>       (github pull #2337 from bp-kelley)
>        - Add a skip_all_if_match option to the FragmentRemover
>       (github pull #2338 from greglandrum)
>        - Dev/general csharp fixes
>       (github pull #2341 from bp-kelley)
>        - Add a read-only Python wrapper for SGroups
>       (github pull #2343 from greglandrum)
>        - Expose RGroupDecomposition to SWIG
>       (github pull #2345 from greglandrum)
>        - update debian build script to python3
>       (github pull #2350 from UnixJunkie)
>        - add GetStereoIsomerCount() function to EnumerateStereoisomers
>       (github pull #2354 from greglandrum)
>        - Update coordgenlibs to v1.2.2
>       (github pull #2355 from ricrogz)
>        - Small fixes to get DLLs to build on Windows
>       (github pull #2356 from ptosco)
>        - Boost deprecation warning
>       (github pull #2357 from d-b-w)
>        - Removes an extra debugging cerr statment
>       (github pull #2360 from d-b-w)
>        - improvements to the Uncharge functionality
>       (github pull #2374 from greglandrum)
>        - Add ExplicitBitVect prop and query
>       (github pull #2384 from bp-kelley)
>        - Allow components of the MolStandardize code to be initialized from streams
>       (github pull #2385 from greglandrum)
>
>     _______________________________________________
>     Rdkit-discuss mailing list
>     Rdk...@li...
>     <mailto:Rdk...@li...>
>     https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
>
>
>
> _______________________________________________
> Rdkit-discuss mailing list
> Rdk...@li...
> https://lists.sourceforge.net/lists/listinfo/rdkit-discuss

[Rdkit-devel] Beta of the 2019.03 release available

[Greg L. <gre...@gm...>]

Dear all,

The beta of the 2019.03 RDKit release has been tagged in github:
https://github.com/rdkit/rdkit/releases/tag/Release_2019_03_1b1

There are a couple more bug fixes and maybe one more feature expected
before the actual release, but I wanted to go ahead and get the beta out
there.

I've done conda builds for Python 3.6 and 3.7 for Windows, Mac, and Linux.
These all use the beta label so that they do not install by default; you'll
need to run "conda install" as follows:

conda install -c rdkit/label/beta rdkit

Be sure to confirm that it's installing the right version when you are
prompted (if there's no build available, it will pick the current
production release instead).

The relevant section of the release notes is below, or you can see a nicely
formatted version here:
https://github.com/rdkit/rdkit/releases/tag/Release_2019_03_1b1

As usual, if you have time to try out the new release I would love
feedback. If nothing major comes up, I plan to do the actual release early
next week.

Best,
-greg

# Release_2019.03.1
(Changes relative to Release_2018.09.1)

## REALLY IMPORTANT ANNOUNCEMENT
- As of this realease (2019.03.1) the RDKit no longer supports Python
2. Please read this rdkit-discuss post to learn what your options are
if you need to keep using Python 2:
  https://www.mail-archive.com/rdk...@li.../msg08354.html

## Backwards incompatible changes
- The fix for github #2245 means that the default behavior of the MaxMinPicker
  is now truly random. If you would like to reproduce the previous behavior,
  provide a seed value of 42.
- The uncharging method in the MolStandardizer now attempts to generate
  canonical results for a given molecule. This may result in different output
  for some molecules.

## Highlights:
- There's now a Japanese translation of large parts of the RDKit documentation
- SGroup data can now be read from and written to Mol/SDF files
- The enhanced stereo handling has been improved: the information is now
  accessible from Python, EnumerateStereoisomers takes advantage of it, and it
  can be read from and written to CXSmiles

## Acknowledgements:
Michael Banck, Francois Berenger, Thomas Blaschke, Brian Cole, Andrew Dalke,
Bakary N'tji Diallo, Guillaume Godin, Jan Holst Jensen, Sunhwan Jo, Brian
Kelley, Petr Kubat, Karl Leswing, Susan Leung, John Mayfield, Adam Moyer, Dan
Nealschneider, Noel O'Boyle, Stephen Roughley, Takayuki Serizawa, Gianluca
Sforna, Ricardo Rodriguez Schmidt, Matt Swain, Paolo Tosco, Ricardo Vianello,
'John-Videogames', 'magattaca', 'msteijaert', 'paconius', 'sirbiscuit'

## Bug Fixes:
  - PgSQL: fix boolean definitions for Postgresql 11
 (github pull #2129 from pkubatrh)
  - update fingerprint tutorial notebook
 (github pull #2130 from greglandrum)
  - Fix typo in RecapHierarchyNode destructor
 (github pull #2137 from iwatobipen)
  - SMARTS roundtrip failure
 (github issue #2142 from mcs07)
  - Error thrown in rdMolStandardize.ChargeParent
 (github issue #2144 from paconius)
  - SMILES parsing inconsistency based on input order
 (github issue #2148 from coleb)
  - MolDraw2D: line width not in python wrapper
 (github issue #2149 from greglandrum)
  - Missing Python API Documentation
 (github issue #2158 from greglandrum)
  - PgSQL: mol_to_svg() changes input molecule.
 (github issue #2174 from janholstjensen)
  - Remove Unicode From AcidBasePair Name
 (github pull #2185 from lilleswing)
  - Inconsistent treatment of `[as]` in SMILES and SMARTS
 (github issue #2197 from greglandrum)
  - RGroupDecomposition fixes, keep userLabels more robust onlyMatchAtRGroups
 (github pull #2202 from bp-kelley)
  - Fix TautomerTransform in operator=
 (github pull #2203 from bp-kelley)
  - testEnumeration hangs/takes where long on 32bit architectures
 (github issue #2209 from mbanck)
  - Silencing some Python 3 warning messages
 (github pull #2223 from coleb)
  - removeHs shouldn't remove atom lists
 (github issue #2224 from rvianello)
  - failure round-tripping mol block with Q atom
 (github issue #2225 from rvianello)
  - problem round-tripping mol files that include bond topology info
 (github issue #2229 from rvianello)
  - aromatic main-group atoms written to SMARTS incorrectly
 (github issue #2237 from greglandrum)
  - findPotentialStereoBonds() stopping too early
 (github issue #2244 from greglandrum)
  - MinMax Diversity picker seeding shows deterministic / non-random behaviour
 (github issue #2245 from sroughley)
  - Fix to serialize binary strings
 (github pull #2264 from bp-kelley)
  - Recognize N in three-membered rings as potentially chiral
 (github issue #2268 from greglandrum)
  - Failure when parsing mol block with M  PXA
 (github issue #2277 from greglandrum)
  - query-query matching failing for atoms constructed from SMARTS
 (github issue #2299 from greglandrum)
  - SMILES parsing fails for dative ring closures
 (github issue #2303 from greglandrum)
  - Missing Dict.h overload: std::string Dict::getVal<std::string>
 (github issue #2308 from greglandrum)
  - fix a problem with the random pickers test
 (github pull #2310 from greglandrum)
  - AddHs() using 3D coordinates with 2D conformations
 (github pull #2328 from greglandrum)
  - change to make the SWIG builds work on windows
 (github pull #2340 from greglandrum)
  - uncharger behaves differently on molecules constructed from mol
blocks and SMILES
 (github issue #2346 from greglandrum)
  - Memory Error When Writing ToBinary With "AllProps"
 (github issue #2352 from atom-moyer)
  - Seg fault on init if RDBASE is not set
 (github issue #2368 from greglandrum)
  - PandasTools.FrameToGridImage() fails with SVG output
 (github issue #2380 from greglandrum)
  - ClusterMols.GetDistanceMatrix throws a type error in Python 3
 (github issue #2387 from John-Videogames)

## New Features and Enhancements:
  - Allow access to Enhanced Stereochemistry information from Python
 (github issue #2108 from d-b-w)
  - Adopt EnumerateStereoisomers to use extended stereo
 (github issue #2109 from greglandrum)
  - Enable ctest -T memcheck
 (github pull #2113 from ricrogz)
  - Support for parsing/writing SGroups in SD Mol files
 (github pull #2138 from ricrogz)
   - Rename the #define _DEBUG to MMPA_DEBUG in mmpa.cpp
 (github pull #2140 from baoilleach)
  - MolDraw2D: line width should be controlled by MolDrawOptions
 (github issue #2151 from greglandrum)
  - Some refactoring of the distance geometry code
 (github pull #2153 from greglandrum)
  - Less warnings
 (github pull #2155 from UnixJunkie)
  - ShapeTverskyIndex python function
 (github pull #2156 from susanhleung)
  - Skip compound if smiles conversion fails
 (github pull #2168 from msteijaert)
  - Fix #2176: InChI functions should return NULL on un-InChI-able
input molecules.
 (github pull #2177 from janholstjensen)
  - Update installation instructions for Linux
 (github pull #2181 from sirbiscuit)
  - Update CMake rules to find external coorgen & maeparser libs
 (github pull #2184 from ricrogz)
  - Update to use the travis Xenial environment
 (github pull #2200 from greglandrum)
  - Do not allow PandasTools to overwrite pandas settings
 (github pull #2206 from sirbiscuit)
  - re-enable (and update) the file parser tests
 (github pull #2208 from greglandrum)
  - Added documentation files written in Japanese into Book directory
 (github pull #2210 from magattaca)
  - Add C++ convenience function for drawing ROMols
 (github issue #2220 from greglandrum)
  - Change boost int types to std types
 (github pull #2233 from bp-kelley)
  - Added exports for SGroup functions
 (github pull #2242 from ricrogz)
  - Use coordMap when starting embedding from random coords
 (github issue #2246 from greglandrum)
  - Improve interactivity of output SVG
 (github pull #2253 from greglandrum)
  - Add options for substructure searching
 (github pull #2254 from greglandrum)
  - keep extra information about bonds from Mol files
 (github pull #2260 from greglandrum)
  - Allow converting mol blocks directly to InChI
 (github pull #2262 from greglandrum)
  - Patch/pains updates
 (github pull #2272 from johnmay)
  - add warning for 2D conformations flagged as 3D
 (github pull #2273 from greglandrum)
  - Store extra CXSMILES data as a property
 (github pull #2281 from ricrogz)
  - Parse enhanced stereo information from CXSMILES
 (github pull #2282 from ricrogz)
  - Robustify parsing of CTABs and SGROUPs
 (github pull #2283 from greglandrum)
  - Write enhanced stereo to cxsmiles
 (github pull #2290 from greglandrum)
  - Allow custom type-handlers in the RDProps interface
 (github pull #2293 from bp-kelley)
  - Add serialization to SubstructLibrary
 (github pull #2295 from bp-kelley)
  - support reading/writing atom props from SD files
 (github pull #2297 from greglandrum)
  - Add test for issue #2285, fix molbundle test
 (github pull #2301 from bp-kelley)
  - Update maeparser & coordgen libraries
 (github pull #2302 from ricrogz)
  - Mem errors clean up
 (github pull #2305 from ricrogz)
  - Add definition of MolFragmentToCXSmiles
 (github pull #2307 from greglandrum)
  - Doc update
 (github pull #2312 from greglandrum)
  - Adds gzstream stream, exposes to swig
 (github pull #2314 from bp-kelley)
  - Remove a bunch of Python2-related warts
 (github pull #2315 from greglandrum)
  - some much-needed optimization work on the new property lists
 (github pull #2317 from greglandrum)
  - Build warnings revisited
 (github pull #2318 from ricrogz)
  - change bogus "3D" to "2D" in a test file
 (github pull #2319 from greglandrum)
  - Allow copying atoms in Python
 (github pull #2322 from d-b-w)
  - fixes an r-group symmetrization problem
 (github pull #2324 from greglandrum)
  - simple docstring fix
 (github pull #2326 from sunhwan)
  - allow using system's catch2 for tests
 (github pull #2327 from giallu)
  - Python wrap DetectAtomStereoChemistry from MolFileStereochem.h
 (github issue #2329 from d-b-w)
  - switch to using cmake to handle the C++ spec
 (github pull #2334 from greglandrum)
  - WIP: optional integration with YAeHMOP
 (github pull #2335 from greglandrum)
  - Exposes substructlibrary to swig
 (github pull #2337 from bp-kelley)
  - Add a skip_all_if_match option to the FragmentRemover
 (github pull #2338 from greglandrum)
  - Dev/general csharp fixes
 (github pull #2341 from bp-kelley)
  - Add a read-only Python wrapper for SGroups
 (github pull #2343 from greglandrum)
  - Expose RGroupDecomposition to SWIG
 (github pull #2345 from greglandrum)
  - update debian build script to python3
 (github pull #2350 from UnixJunkie)
  - add GetStereoIsomerCount() function to EnumerateStereoisomers
 (github pull #2354 from greglandrum)
  - Update coordgenlibs to v1.2.2
 (github pull #2355 from ricrogz)
  - Small fixes to get DLLs to build on Windows
 (github pull #2356 from ptosco)
  - Boost deprecation warning
 (github pull #2357 from d-b-w)
  - Removes an extra debugging cerr statment
 (github pull #2360 from d-b-w)
  - improvements to the Uncharge functionality
 (github pull #2374 from greglandrum)
  - Add ExplicitBitVect prop and query
 (github pull #2384 from bp-kelley)
  - Allow components of the MolStandardize code to be initialized from streams
 (github pull #2385 from greglandrum)

Re: [Rdkit-devel] smiles parsing - rings

[Jonas B. <dr....@gm...>]

Hi Andrew,
Good to hear from you, and thanks for the info. I might go for a
quick-n-dirty fix then. The reason I am asking is that I am rewriting an
OEChem script to RDkit to share with a collaborator, and this way of doing
it (with parentheses) works fine in OEChem, with the OESmilesToMol
function.
Cheers
Jonas

tis 20 nov. 2018 kl. 17:04 skrev Andrew Dalke <da...@da...>:

> Hi Jonas,
>
> > On Nov 20, 2018, at 16:36, Jonas Boström <dr....@gm...>
> wrote:
> > not_ok_smiles = "C%43CCC%42.CC%42.CC(%43)C"
>
> The error message from RDKit is the not-very-helpful:
>
>    SMILES Parse Error: syntax error for input: 'C%43CCC%42.CC%42.CC(%43)C'
>
> If you try out my 'smiview' (available via 'pip install smiview' for
> Python):
>
>   % python smiview.py 'C%43CCC%42.CC%42.CC(%43)C'
>   Cannot parse --smiles: Unexpected term
>     C%43CCC%42.CC%42.CC(%43)C
>                         ^ Tokenizing stopped here
>   An open branch must be followed by an atom, bond, or dot.
>
> In Daylight and OpenSMILES, the ring closures must occur immediately after
> the atom symbol. More specifically, quoting
> http://www.daylight.com/dayhtml/doc/theory/theory.smiles.html :
>
>   The bonds are numbered in any order, designating ring opening (or ring
> closure)
>   bonds by a digit immediately following the atomic symbol at each ring
> closure.
>
>
> RDKit also allows ring closures after a ')', which is connected to the
> atom symbol before the matching '('.
>
> Cheers,
>
>                                 Andrew
>                                 da...@da...
>
>
>
>
> _______________________________________________
> Rdkit-devel mailing list
> Rdk...@li...
> https://lists.sourceforge.net/lists/listinfo/rdkit-devel
>

Re: [Rdkit-devel] smiles parsing - rings

[Andrew D. <da...@da...>]

Hi Jonas,

> On Nov 20, 2018, at 16:36, Jonas Boström <dr....@gm...> wrote:
> not_ok_smiles = "C%43CCC%42.CC%42.CC(%43)C"

The error message from RDKit is the not-very-helpful:

   SMILES Parse Error: syntax error for input: 'C%43CCC%42.CC%42.CC(%43)C'

If you try out my 'smiview' (available via 'pip install smiview' for Python):

  % python smiview.py 'C%43CCC%42.CC%42.CC(%43)C'
  Cannot parse --smiles: Unexpected term
    C%43CCC%42.CC%42.CC(%43)C
                        ^ Tokenizing stopped here
  An open branch must be followed by an atom, bond, or dot.

In Daylight and OpenSMILES, the ring closures must occur immediately after the atom symbol. More specifically, quoting http://www.daylight.com/dayhtml/doc/theory/theory.smiles.html :

  The bonds are numbered in any order, designating ring opening (or ring closure)
  bonds by a digit immediately following the atomic symbol at each ring closure.


RDKit also allows ring closures after a ')', which is connected to the atom symbol before the matching '('.

Cheers,

				Andrew
				da...@da...

[Rdkit-devel] smiles parsing - rings

[Jonas B. <dr....@gm...>]

Hello,
I am trying to get my head around how smiles parsing in RDkit works,
and wonder if anyone can explain why the "ok_smiles" is processed ok,
wheras the "not_ok_smiles", with the extra parenthesis, isn't?

ok_smiles = "C%43CCC%42.CC%42.CC%43C"
ok_mol = Chem.MolFromSmiles(ok_smiles)
print ("ok smiles", Chem.MolToSmiles(ok_mol))

not_ok_smiles = "C%43CCC%42.CC%42.CC(%43)C"
not_ok_mol = Chem.MolFromSmiles(not_ok_smiles)
print ("not ok smiles", Chem

Thanks
Jonas
I am running the below

RDKit version 2018.03.2

Python 2.7.15 |Anaconda, Inc.| (default, May  1 2018, 23:32:55)

[Rdkit-devel] 2018.09.1 RDKit Release

[Greg L. <gre...@gm...>]

I'm pleased to announce that the next version of the RDKit - 2018.09 - is
released. The release notes are below.

The release files are on the github release page:
https://github.com/rdkit/rdkit/releases/tag/Release_2018_09_1

Binaries have been uploaded to anaconda.org (https://anaconda.org/rdkit).
The available conda binaries for this release are:
Linux 64bit: python 2.7, 3.6
Mac OS 64bit: python 2.7, 3.6
Windows 64bit: python 3.6

I hope to be able to do python 3.7 binaries sometime in the not-too-distant
future.

The missing binary for python 2.7 on windows is explained here:
https://www.mail-archive.com/rdk...@li.../msg07850.html

Please note that this is the last feature release of the RDKit which will
support Python 2.7. More details on this here:
https://www.mail-archive.com/rdk...@li.../msg08354.html

I've also put conda builds of the PostgreSQL cartridge up for Linux and the
Mac.

The online version of the documentation at rdkit.org (
http://rdkit.org/docs/index.html) has been updated. Note that because of
the move away from epydoc, the structure of the documentation has changed a
bit and old URLs may not work.

Some things that will be finished over the next couple of days:
- The conda build scripts will be updated to reflect the new version
- The homebrew script

Thanks to everyone who submitted code, bug reports, and suggestions for
this release!

Please let me know if you find any problems with the release or have
suggestions for the next one, which is scheduled for April 2019.

Best Regards,
-greg


# Release_2018.09.1
(Changes relative to Release_2018.03.1)

## Highlights:
- This release includes two contributions from the Google Summer of Code:
   - A new generalized fingerprint generator.
   - An integration/port of MolVS to the core RDKit.
  The API on both of these may change a bit with future releases.
- The rdkit.Chem.Draw module now includes functions for depicting
fingerprint
  bits. Morgan and RDKit fingerprints are currently supported.

## Deprecations
- As part of the changes and cleaning up done for #1836 many of the
`#defines`
  used in the RDKit have been renamed.
    - `USE_BUILTIN_POPCOUNT` -> `RDK_OPTIMIZE_NATIVE`
    - `RDK_THREADSAFE_SSS` -> `RDK_BUILD_THREADSAFE_SSS`
    - `BUILD_COORDGEN_SUPPORT` -> `RDK_BUILD_COORDGEN_SUPPORT`
    - `BUILD_AVALON_SUPPORT` -> `RDK_BUILD_AVALON_SUPPORT`
    - `BUILD_INCHI_SUPPORT` -> `RDK_BUILD_INCHI_SUPPORT`
    - `BUILD_SLN_SUPPORT` -> `RDK_BUILD_SLN_SUPPORT`
    - `RDK_CAIRO_BUILD` -> `RDK_BUILD_CAIRO_SUPPORT`

## Documentation updates
We have moved to using Sphinx's autodoc to create the Python API
documentation.
`epydoc`, the tool we used to use, is no longer actively developed and only
supports
Python2. There will undoubtedly be problems associated with the change; if
you notice
anything missing in the documetation or something that's really badly
formatted,
please either let us know or submit a PR with a fix.

## Backwards incompatible changes
This release includes a set of changes to make the default arguments to
common
functions less error prone (github #1679).
- GetAtomSmiles() now generates isomeric SMILES by default.
- The ringMatchesRingOnly option to the FindMCS() function now applies to
  atom-atom matches as well as bond-bond matches.
- The Python functions EmbedMolecule() and EmbedMultipleConfs() now use the
  ETKDG algorithm by default instead of standard distance geometry.

## Acknowledgements:
Boran Adas, Francois Berenger, Thomas Blaschke,  Brian Cole, Andrew Dalke,
Guillaume Godin,
Brice Hoff, Brian Kelley, Karl Leswing, Susan Leung, Pat Lorton, Josh
Meyers, Hirotomo Moriwaki,
Dan Nealschneider, Noel O'Boyle, Pavel Raiskup, Sereina Riniker, Ricardo
Rodriguez Schmidt,
Stephen Roughley, Roger Sayle, Takayuki Serizawa, Rim Shayakhmetov, Gregory
Simm, Jon Sorenson,
Matt Swain, Kiran Telukunta, Paulo Tosco, Alain Vaucher, Maciej
Wójcikowski, '0xDECAFC0FFEE',
'jaechanglim', 'paconius'

## Contrib updates:
- The FastCluster code has been updated.

## New Features and Enhancements:
  - expose MolChemicalFeature.{get,set}ActiveConformer()  to python
 (github issue #1636 from greglandrum)
  - added Scripts/create_deb_packages.sh
 (github pull #1655 from UnixJunkie)
  - Start to use Catch2 for new tests
 (github pull #1732 from greglandrum)
  - Switch DbCLI scripts from optparse to argparse
 (github issue #1778 from greglandrum)
  - Add EEM partial charges
 (github pull #1828 from greglandrum)
  - Add header file providing access to RDKit compile time flags
 (github issue #1836 from greglandrum)
  - add control over the coordinate precision to coordgen
 (github pull #1847 from greglandrum)
  - Add Properties interface to ChemicalReactions
 (github pull #1848 from greglandrum)
  - Switch Python API documentation generation over to Sphinx autodoc
 (github pull #1849 from greglandrum)
  - expose MolOps::assignStereochemistryFrom3D() to Python
 (github issue #1850 from greglandrum)
  - bivariate_normal deprecation in mlab.py of matplotlib
 (github issue #1851 from telukir)
  - Expose minAtomRingSize() and minBondRingSize() to Python wrappers
 (github pull #1859 from mwojcikowski)
  - enable building DLLs on Windows
 (github pull #1861 from ptosco)
  - Fix compatibility with Boost 1.67+
 (github pull #1864 from mcs07)
  - Don't manually set RDConfig paths in conda env
 (github pull #1865 from mcs07)
  - Make svg xmlns prefix use more consistent
 (github pull #1866 from mcs07)
  - Add custom 3D Descriptors
 (github pull #1867 from greglandrum)
  - Add initial Maestro format Supplier using maeparser library
 (github pull #1872 from lorton)
  - add queryAtomNonHydrogenDegree() query operation
 (github issue #1873 from greglandrum)
  - Add an auto-populated file with cmake config options
 (github pull #1874 from greglandrum)
  - Custom property VSA
 (github pull #1884 from sriniker)
  - Swap maeparser and coordgen dependencies to use releases
 (github issue #1887 from greglandrum)
  - expose MolDraw2DSVG.tagAtoms() to python
 (github pull #1897 from greglandrum)
  - allow the cleanup step of Mol2 parsing to be disabled
 (github pull #1898 from greglandrum)
  - Allow Atom.GetAtomSmarts() to return isomeric SMILES
 (github pull #1902 from greglandrum)
   - Allow coordgen and maeparser to be built as static libraries
 (github pull #1909 from ptosco)
  - Support reaction_to_svg() in cartridge
 (github issue #1916 from greglandrum)
  - Addresses several minor warning messages during the build
 (github pull #1935 from d-b-w)
  - Some optimization of the queries constructed from SMARTS
 (github pull #1940 from greglandrum)
  - Add ring/chain match constraints options to AdjustQueryParameters()
 (github issue #1943 from greglandrum)
  - roc calculation naming problem
 (github pull #1975 from 0xDECAFC0FFEE)
  - Fingerprinting functions should call assignStereochemistry() when
necessary
 (github issue #1993 from bricehoff)
  - Dev/GSOC2018_MolVS_Integration
 (github pull #2002 from susanhleung)
  - GSoC 2018 - Fingerprints
 (github pull #2005 from Boranadas)
  - port fingerprint bit rendering code from CheTo
 (github pull #2008 from greglandrum)
  - PgSQL: add support for PostgreSQL_CONFIG cmake var
 (github pull #2014 from praiskup)
  - Add missing boost header for v1.64
 (github pull #2016 from gncs)
  - Enhanced stereo read/write support in SDF files.
 (github pull #2022 from d-b-w)
  - IPythonConsole hooks should copy the original docstring
 (github issue #2025 from adalke)
  - Allow dumping interchange information into SVG files
 (github pull #2030 from greglandrum)
  - MCS: add test for ring--non-ring matches at the atom level
 (github issue #2034 from greglandrum)
  - Ability to generate a list of possible smiles representation for a
given molecule
 (github issue #2042 from thegodone)
  - add scoring test (relevant to #1975)
 (github pull #2064 from greglandrum)
  - actually run the SmilesWriterNoNames() test
 (github pull #2067 from greglandrum)
  - Add a default for maximum products generated by a reaction (maxProduc…
 (github pull #2069 from bp-kelley)
  - Add user-defined literals for parsing SMILES and SMARTS
 (github pull #2070 from greglandrum)
  - move rdInfoLog to stderr
 (github pull #2073 from greglandrum)
  - add confId argument to MolChemicalFeatureFactor::getFeaturesForMol()
 (github issue #2077 from greglandrum)
  - Expose a CMake flag to build RDKit with -rpath
 (github pull #2084 from coleb)
  - Dev/expose setquery to python
 (github pull #2088 from bp-kelley)
  - Updated .gitignore with files generated outside of build directory.
 (github pull #2095 from ricrogz)
  - Address compile warnings & trivial improvements
 (github pull #2097 from ricrogz)
  - Coordgen: stop printing Templates location
 (github pull #2102 from greglandrum)
  - Update Docs For CalcBEDROC
 (github pull #2103 from lilleswing)

## Bug Fixes:
  - Cannot find rings for ChEBI 50252
 (github issue #299 from greglandrum)
  - Feature request: #defines to test RDKit version
 (github issue #1454 from baoilleach)
  - Atoms in residue read from pdb lose their AtomPDBResidueInfo after
reaction (Bug)
 (github issue #1632 from hjuinj)
  - SMARTS parse failure for some queries involving Hs
 (github issue #1719 from greglandrum)
  - Conformer indexing bug in RDFreeSASA.cpp?
 (github issue #1786 from paconius)
  - allow libs to not be installed
 (github pull #1832 from greglandrum)
  - RWMol::addAtom(Atom,bool) missing from Java wrappers
 (github issue #1837 from greglandrum)
  - RWMol::clear now calls ROMol::initMol
 (github pull #1844 from bp-kelley)
  - Allow types.h to be included in applications that use
/D_USE_MATH_DEFINES
 (github pull #1846 from d-b-w)
  - Fixes failing Python tests on Windows
 (github pull #1855 from ptosco)
  - Allow building on cygwin using -std=c++11
 (github pull #1856 from greglandrum)
  - Stop using the cmake Boost:: targets
 (github pull #1858 from greglandrum)
  - R-group Decomposition: allow H replacements when matchOnlyAtRgroups is
set
 (github pull #1871 from bp-kelley)
  - Mark cartridge functions as being parallel safe
 (github issue #1886 from greglandrum)
  - Fixes locale handling on Windows
 (github pull #1892 from ptosco)
  - get the pandas tests working with pandas 0.23
 (github pull #1896 from greglandrum)
  - "make install" appears to miss RDBoost/export.h
 (github issue #1903 from baoilleach)
  - Fix curl fallback for downloading files
 (github pull #1904 from d-b-w)
  - Bond stereo information not output to SMARTS
 (github issue #1906 from greglandrum)
  - Library .so names missing RDKit?
 (github issue #1913 from baoilleach)
  - Negated atom number queries in SMARTS should not set atomic number of
query atom
 (github issue #1920 from greglandrum)
  - memory leak in Get3DDistanceMatrix
 (github issue #1924 from jaechanglim)
  - Atom with bond to itself is accepted by the SMILES parser.
 (github issue #1925 from tblaschke)
  - Possibly incorrect aromatic SMILES generated for structure
 (github issue #1928 from baoilleach)
  - Using the coordgen library seems to cause a seg fault
 (github issue #1929 from JoshuaMeyers)
  - Aromaticity failure in 7-ring with charged radical carbon
 (github issue #1936 from bp-kelley)
  - Fix embarassing bug, check the counter each iteration
 (github pull #1939 from bp-kelley)
  - RuntimeError when importing rdkit.Chem.Descriptors with Python 3.7
 (github issue #1948 from drkeoni)
  - Query features in products of rxn files not properly handled
 (github issue #1950 from greglandrum)
  - ReactionToSmarts broken for multi-component templates
 (github issue #1955 from bp-kelley)
  - update knime urls in overview.md
 (github pull #1966 from greglandrum)
  - CXN extended SMILES labels are not applied to the correct atom in
SMILES with explicit H
 (github issue #1968 from baoilleach)
  - MolFromSmarts MolToSmarts fails to roundtrip on patterns with chirality
 (github issue #1985 from bp-kelley)
  - QueryAtoms with atom list queries should not have the atomic number set
 (github issue #1988 from greglandrum)
  - RemoveHs not properly updating double bond stereoatoms
 (github issue #1990 from shayakhmetov)
  - Error while parsing empty atom list in Mol files.
 (github issue #2000 from drkeoni)
  - Cleanup step of sanitization sometimes sets undesired formal charges
 (github issue #2020 from avaucher)
  - GetBondSmiles() returns nothing for wedged bonds when allBondsExplicit
is true
 (github issue #2029 from greglandrum)
  - PMIs and NPRs return same value between different conformer
 (github issue #2037 from philopon)
  - Failure to parse V3K mol file with bonds to multi-center linkage points
 (github issue #2040 from greglandrum)
  - patch a memory allocation problem in the maeparser v1.0.0
 (github pull #2044 from greglandrum)
  - CIPRank values from JSONDataToMols are not unsigned
 (github issue #2046 from greglandrum)
  - switch to v1.0.1 of the maeparser
 (github pull #2048 from greglandrum)
  - Update fastcluster code
 (github pull #2051 from greglandrum)
  - Fix memory leak in Dict operator=
 (github pull #2061 from bp-kelley)
  - Atom mapping lost after call to MergeQueryHs()
 (github issue #2062 from greglandrum)
  - Drawing racemic bond stereo as crossed bonds should be the default
 (github issue #2063 from coleb)
  - Moved test.h from RDBoost to RDGeneral for consistency with export.h
 (github pull #2074 from ptosco)
  - CTABs behave differently than before regarding stereo
 (github issue #2082 from bp-kelley)
  - BitInfo not complete for RDKFingerprint
 (github issue #2115 from greglandrum)

## Removed code:
- Remove the deprecated MolDrawing code
 (github pull #2111 from greglandrum)

[Rdkit-devel] [Question] Ok to switch to conda-forge for RDKit builds?

[Greg L. <gre...@gm...>]

Dear all,

TL;DR: Now that RDKit builds are available on conda-forge, I would like to
stop doing builds on the rdkit conda channel. I'm looking for feedback
about that here.

Thanks to the persistence of Peter St John and Matt Swain, the RDKit is now
available on conda-forge:
https://anaconda.org/conda-forge/rdkit/files
If you take a look, you'll find builds for Mac, Linux, and Windows together
with multiple versions of Python. The build/deployment steps are all
automated. It's really cool!

Given that these packages now exist, I don't see much value in continuing
to also host standard RDKit builds in the rdkit conda channel. So I would
propose that we switch to using conda-forge as the primary channel for
doing RDKit builds/distributions. Having the automated process available
would save me a lot of time and remove the chance for me to make a mistake
while doing the builds.[1]
If we go this way, I will update the documentation to point to conda-forge
and will only do the occasional nightly or beta build in the rdkit conda
channel.

Does anyone have any objections to this or see any problems that it might
cause?

Many thanks to Matt and Peter for making the conda-forge builds happen!

-greg
[1] As fun as these are, I think I can live without them. ;-)

[Rdkit-devel] Beta of the 2018.09 release available

[Greg L. <gre...@gm...>]

Dear all,

The beta of the 2018.09 RDKit release has been tagged in github:
https://github.com/rdkit/rdkit/releases/tag/Release_2018_09_1b1

I will try to get some conda builds for python 3.6 up over the next day or
so. These will use the beta label so that they do not install by default;
you'll need to run "conda install" as follows:

conda install -c rdkit/label/beta rdkit python=3.6

Be sure to confirm that it's installing the right version when you are
prompted (if there's no build available, it will pick the current
production release instead).

You can check to see if a build is available for your platform/python
version here:
*https://anaconda.org/rdkit/rdkit/files?version=2018.09.1b1.0
<https://anaconda.org/rdkit/rdkit/files?version=2018.09.1b1.0>*

The relevant section of the release notes is below.

As usual, if you have time to try out the new release I would love
feedback. If nothing major comes up, I plan to do the actual release
sometime next week.

Best,
-greg

# Release_2018.09.1
(Changes relative to Release_2018.03.1)

## Deprecations
- As part of the changes and cleaning up done for #1836 many of the
`#defines`
  used in the RDKit have been renamed.
    - `USE_BUILTIN_POPCOUNT` -> `RDK_OPTIMIZE_NATIVE`
    - `RDK_THREADSAFE_SSS` -> `RDK_BUILD_THREADSAFE_SSS`
    - `BUILD_COORDGEN_SUPPORT` -> `RDK_BUILD_COORDGEN_SUPPORT`
    - `BUILD_AVALON_SUPPORT` -> `RDK_BUILD_AVALON_SUPPORT`
    - `BUILD_INCHI_SUPPORT` -> `RDK_BUILD_INCHI_SUPPORT`
    - `BUILD_SLN_SUPPORT` -> `RDK_BUILD_SLN_SUPPORT`
    - `RDK_CAIRO_BUILD` -> `RDK_BUILD_CAIRO_SUPPORT`

## Backwards incompatible changes
This release includes a set of changes to make the default arguments to
common
functions less error prone (github #1679).
- GetAtomSmiles() now generates isomeric SMILES by default.
- The ringMatchesRingOnly option to the FindMCS() function now applies to
  atom-atom matches as well as bond-bond matches.
- The Python functions EmbedMolecule() and EmbedMultipleConfs() now use the
  ETKDG algorithm by default instead of standard distance geometry.

## Highlights:
- This release includes two contributions from the Google Summer of Code:
   - A new generalized fingerprint generator.
   - An integration/port of MolVS to the core RDKit.
  The API on both of these may change a bit with future releases.
- The rdkit.Chem.Draw module now includes functions for depicting
fingerprint
  bits. Morgan and RDKit fingerprints are currently supported.

## Acknowledgements:
Boran Adas, Francois Berenger, Thomas Blaschke,  Brian Cole, Andrew Dalke,
Guillaume Godin, Brice Hoff, Brian Kelley, Karl Leswing, Susan Leung, Pat
Lorton, Josh Meyers, Hirotomo Moriwaki, Dan Nealschneider, Noel O'Boyle,
Pavel Raiskup, Sereina Riniker, Ricardo Rodriguez Schmidt, Stephen
Roughley, Roger Sayle, Rim Shayakhmetov, Gregory Simm, Jon Sorenson, Matt
Swain, Kiran Telukunta, Paulo Tosco, Alain Vaucher, Maciej Wójcikowski,
'0xDECAFC0FFEE', 'jaechanglim', 'paconius'

## Contrib updates:
- The FastCluster code has been updated.

## New Features and Enhancements:
  - expose MolChemicalFeature.{get,set}ActiveConformer()  to python
 (github issue #1636 from greglandrum)
  - added Scripts/create_deb_packages.sh
 (github pull #1655 from UnixJunkie)
  - Start to use Catch2 for new tests
 (github pull #1732 from greglandrum)
  - Switch DbCLI scripts from optparse to argparse
 (github issue #1778 from greglandrum)
  - Add EEM partial charges
 (github pull #1828 from greglandrum)
  - Add header file providing access to RDKit compile time flags
 (github issue #1836 from greglandrum)
  - add control over the coordinate precision to coordgen
 (github pull #1847 from greglandrum)
  - Add Properties interface to ChemicalReactions
 (github pull #1848 from greglandrum)
  - Switch Python API documentation generation over to Sphinx autodoc
 (github pull #1849 from greglandrum)
  - expose MolOps::assignStereochemistryFrom3D() to Python
 (github issue #1850 from greglandrum)
  - bivariate_normal deprecation in mlab.py of matplotlib
 (github issue #1851 from telukir)
  - Expose minAtomRingSize() and minBondRingSize() to Python wrappers
 (github pull #1859 from mwojcikowski)
  - enable building DLLs on Windows
 (github pull #1861 from ptosco)
  - Fix compatibility with Boost 1.67+
 (github pull #1864 from mcs07)
  - Don't manually set RDConfig paths in conda env
 (github pull #1865 from mcs07)
  - Make svg xmlns prefix use more consistent
 (github pull #1866 from mcs07)
  - Add custom 3D Descriptors
 (github pull #1867 from greglandrum)
  - Add initial Maestro format Supplier using maeparser library
 (github pull #1872 from lorton)
  - add queryAtomNonHydrogenDegree() query operation
 (github issue #1873 from greglandrum)
  - Add an auto-populated file with cmake config options
 (github pull #1874 from greglandrum)
  - Custom property VSA
 (github pull #1884 from sriniker)
  - Swap maeparser and coordgen dependencies to use releases
 (github issue #1887 from greglandrum)
  - expose MolDraw2DSVG.tagAtoms() to python
 (github pull #1897 from greglandrum)
  - allow the cleanup step of Mol2 parsing to be disabled
 (github pull #1898 from greglandrum)
  - Allow Atom.GetAtomSmarts() to return isomeric SMILES
 (github pull #1902 from greglandrum)
   - Allow coordgen and maeparser to be built as static libraries
 (github pull #1909 from ptosco)
  - Support reaction_to_svg() in cartridge
 (github issue #1916 from greglandrum)
  - Addresses several minor warning messages during the build
 (github pull #1935 from d-b-w)
  - Some optimization of the queries constructed from SMARTS
 (github pull #1940 from greglandrum)
  - Add ring/chain match constraints options to AdjustQueryParameters()
 (github issue #1943 from greglandrum)
  - roc calculation naming problem
 (github pull #1975 from 0xDECAFC0FFEE)
  - Fingerprinting functions should call assignStereochemistry() when
necessary
 (github issue #1993 from bricehoff)
  - Dev/GSOC2018_MolVS_Integration
 (github pull #2002 from susanhleung)
  - GSoC 2018 - Fingerprints
 (github pull #2005 from Boranadas)
  - port fingerprint bit rendering code from CheTo
 (github pull #2008 from greglandrum)
  - PgSQL: add support for PostgreSQL_CONFIG cmake var
 (github pull #2014 from praiskup)
  - Add missing boost header for v1.64
 (github pull #2016 from gncs)
  - Enhanced stereo read/write support in SDF files.
 (github pull #2022 from d-b-w)
  - IPythonConsole hooks should copy the original docstring
 (github issue #2025 from adalke)
  - Allow dumping interchange information into SVG files
 (github pull #2030 from greglandrum)
  - MCS: add test for ring--non-ring matches at the atom level
 (github issue #2034 from greglandrum)
  - add scoring test (relevant to #1975)
 (github pull #2064 from greglandrum)
  - actually run the SmilesWriterNoNames() test
 (github pull #2067 from greglandrum)
  - Add a default for maximum products generated by a reaction (maxProduc…
 (github pull #2069 from bp-kelley)
  - Add user-defined literals for parsing SMILES and SMARTS
 (github pull #2070 from greglandrum)
  - move rdInfoLog to stderr
 (github pull #2073 from greglandrum)
  - add confId argument to MolChemicalFeatureFactor::getFeaturesForMol()
 (github issue #2077 from greglandrum)
  - Expose a CMake flag to build RDKit with -rpath
 (github pull #2084 from coleb)
  - Dev/expose setquery to python
 (github pull #2088 from bp-kelley)
  - Updated .gitignore with files generated outside of build directory.
 (github pull #2095 from ricrogz)
  - Address compile warnings & trivial improvements
 (github pull #2097 from ricrogz)
  - Coordgen: stop printing Templates location
 (github pull #2102 from greglandrum)
  - Update Docs For CalcBEDROC
 (github pull #2103 from lilleswing)

## Bug Fixes:
  - Cannot find rings for ChEBI 50252
 (github issue #299 from greglandrum)
  - Feature request: #defines to test RDKit version
 (github issue #1454 from baoilleach)
  - Atoms in residue read from pdb lose their AtomPDBResidueInfo after
reaction (Bug)
 (github issue #1632 from hjuinj)
  - SMARTS parse failure for some queries involving Hs
 (github issue #1719 from greglandrum)
  - Conformer indexing bug in RDFreeSASA.cpp?
 (github issue #1786 from paconius)
  - allow libs to not be installed
 (github pull #1832 from greglandrum)
  - RWMol::addAtom(Atom,bool) missing from Java wrappers
 (github issue #1837 from greglandrum)
  - RWMol::clear now calls ROMol::initMol
 (github pull #1844 from bp-kelley)
  - Allow types.h to be included in applications that use
/D_USE_MATH_DEFINES
 (github pull #1846 from d-b-w)
  - Fixes failing Python tests on Windows
 (github pull #1855 from ptosco)
  - Allow building on cygwin using -std=c++11
 (github pull #1856 from greglandrum)
  - Stop using the cmake Boost:: targets
 (github pull #1858 from greglandrum)
  - R-group Decomposition: allow H replacements when matchOnlyAtRgroups is
set
 (github pull #1871 from bp-kelley)
  - Mark cartridge functions as being parallel safe
 (github issue #1886 from greglandrum)
  - Fixes locale handling on Windows
 (github pull #1892 from ptosco)
  - get the pandas tests working with pandas 0.23
 (github pull #1896 from greglandrum)
  - "make install" appears to miss RDBoost/export.h
 (github issue #1903 from baoilleach)
  - Fix curl fallback for downloading files
 (github pull #1904 from d-b-w)
  - Bond stereo information not output to SMARTS
 (github issue #1906 from greglandrum)
  - Library .so names missing RDKit?
 (github issue #1913 from baoilleach)
  - Negated atom number queries in SMARTS should not set atomic number of
query atom
 (github issue #1920 from greglandrum)
  - memory leak in Get3DDistanceMatrix
 (github issue #1924 from jaechanglim)
  - Atom with bond to itself is accepted by the SMILES parser.
 (github issue #1925 from tblaschke)
  - Possibly incorrect aromatic SMILES generated for structure
 (github issue #1928 from baoilleach)
  - Using the coordgen library seems to cause a seg fault
 (github issue #1929 from JoshuaMeyers)
  - Aromaticity failure in 7-ring with charged radical carbon
 (github issue #1936 from bp-kelley)
  - Fix embarassing bug, check the counter each iteration
 (github pull #1939 from bp-kelley)
  - RuntimeError when importing rdkit.Chem.Descriptors with Python 3.7
 (github issue #1948 from drkeoni)
  - Query features in products of rxn files not properly handled
 (github issue #1950 from greglandrum)
  - ReactionToSmarts broken for multi-component templates
 (github issue #1955 from bp-kelley)
  - update knime urls in overview.md
 (github pull #1966 from greglandrum)
  - CXN extended SMILES labels are not applied to the correct atom in
SMILES with explicit H
 (github issue #1968 from baoilleach)
  - MolFromSmarts MolToSmarts fails to roundtrip on patterns with chirality
 (github issue #1985 from bp-kelley)
  - QueryAtoms with atom list queries should not have the atomic number set
 (github issue #1988 from greglandrum)
  - RemoveHs not properly updating double bond stereoatoms
 (github issue #1990 from shayakhmetov)
  - Error while parsing empty atom list in Mol files.
 (github issue #2000 from drkeoni)
  - Cleanup step of sanitization sometimes sets undesired formal charges
 (github issue #2020 from avaucher)
  - GetBondSmiles() returns nothing for wedged bonds when allBondsExplicit
is true
 (github issue #2029 from greglandrum)
  - PMIs and NPRs return same value between different conformer
 (github issue #2037 from philopon)
  - Failure to parse V3K mol file with bonds to multi-center linkage points
 (github issue #2040 from greglandrum)
  - patch a memory allocation problem in the maeparser v1.0.0
 (github pull #2044 from greglandrum)
  - CIPRank values from JSONDataToMols are not unsigned
 (github issue #2046 from greglandrum)
  - switch to v1.0.1 of the maeparser
 (github pull #2048 from greglandrum)
  - Update fastcluster code
 (github pull #2051 from greglandrum)
  - Fix memory leak in Dict operator=
 (github pull #2061 from bp-kelley)
  - Atom mapping lost after call to MergeQueryHs()
 (github issue #2062 from greglandrum)
  - Drawing racemic bond stereo as crossed bonds should be the default
 (github issue #2063 from coleb)
  - Moved test.h from RDBoost to RDGeneral for consistency with export.h
 (github pull #2074 from ptosco)
  - CTABs behave differently than before regarding stereo
 (github issue #2082 from bp-kelley)

## Removed code:
- Remove the deprecated MolDrawing code
 (github pull #2111 from greglandrum)

Re: [Rdkit-devel] The RDKit Travis builds

[Greg L. <gre...@gm...>]

The good news is that I didn't end up needing to do anything: after the
first build failed strangely the others worked normally. Yay! So we're now
migrated to travis-ci.com:
https://travis-ci.com/rdkit/rdkit

The bad news is that the intermittent, non-reproducible build failures seem
to still be an issue. Grrrrr

-greg

[Rdkit-devel] The RDKit Travis builds

[Greg L. <gre...@gm...>]

Just a quick FYI that I asked the folks at Travis-CI to switch us over to
their newer setup (Travis-ci.com instead of Travis-ci.org) today.[1] This
has broken the build, as one would expect. :-S

I will spend some time on it over the next day or so to get things working
again.

-greg
[1] there were intermittent and non-reproducible build failures on the old
system and I am hoping this will help. It needs to happen eventually anyway.