Greg Landrum

Old stereochemistry perception discards ring stereo in spiro systems

5

**Describe the bug** In some cases `assignStereochemistry()` discards stereo markers for molecules which have ring stereo and spiro linkages. It looks like the new stereo code doesn't have this problem...

**Describe the bug** No aromaticity is found in this molecule, `C1=CC2=CC3=CC=C4C=C1C2C34` :  The outer envelope (highlighted in the image) should be aromatic. What's happening is that none of the...

Can't roundtrip CHEMBL4080644 through SMILES

4

**Describe the bug** The SMILES generated from the molblock for CHEMBL4080644 cannot be sanitized: ``` In [2]: chembl4080644 = """ RDKit 2D 22 25 0 0 0 0 0 0...

New version of AssignStereochemistry

1

We added support for new code to find potential stereocenters in #3324 for the 2020.09 release. The idea was to eventually be able to replace `assignStereochemistry()` with something more flexible....

Here's a compact demonstration: ``` In [12]: m = Chem.MolFromSmiles("c1ccc([C@@H](CC2=NCCN2)c2ccccn2)cc1") In [13]: nm = Chem.Mol(m) In [14]: Chem.Kekulize(nm, True) In [15]: Chem.MolToSmiles(enum.Canonicalize(m)) Out[15]: 'c1ccc([C@@H](CC2=NCCN2)c2ccccn2)cc1' In [16]: Chem.MolToSmiles(enum.Canonicalize(nm)) Out[16]: 'c1ccc(C(CC2=NCCN2)c2ccccn2)cc1' ```...

Add 3D embedding to the cartridge

2

Including appropriate options, of course.

[WIP] simplify valence calculations

3

The current code for calculating the explicit/implicit valence of an atom has evolved over time and has become byzantine and hard to understand/explain. This changes that by simplifying the way...

Potentially a bug in reactions on molecules with more than one fragment

1

**Description:** If you apply a reaction to a molecule with more than one fragment, any fragment that has no atoms involved in the reaction will not be transferred to the...

[while doing some large-scale regression testing I found a problem with this. Will fix it, remove the draft marker, and delete this remark] This is one of those "can't believe...

There's some order-dependence problem in the calculation of canonical atom ranks here: ``` from rdkit.Chem import Draw m2a = Chem.MolFromSmiles('C([C@H](C)C1)[C@]12CCN2') m2b = Chem.MolFromSmiles('C[C@H](C1)C[C@]12CCN2') aranks = Chem.CanonicalRankAtoms(m2a,breakTies=False) branks = Chem.CanonicalRankAtoms(m2b,breakTies=False) for...