[Rdkit-discuss] How to determine if atoms are part of the same ring?

2017-02-08 Thread Markus Metz
Dear RDKit community: I am facing the situation that I would like to attach indices to atoms in a molecule defined by a substructure. The substitution pattern is of importance. Therefore I needed to include the wildcards atoms in my substructure. If I get the atom indices of this substructure,

Re: [Rdkit-discuss] How to determine if atoms are part of the same ring?

2017-02-08 Thread Andrew Dalke
On Feb 8, 2017, at 19:22, Markus Metz wrote: > The question to you is: Is there another more elegant way of doing it? May be > I missed something from the python API? I don't quite follow what you are looking for, though I have managed to condense your code somewhat, into:

Re: [Rdkit-discuss] Stereochemistry issue for spirocycles/pseudochiral centres(?)

2017-02-08 Thread James Davidson
Hi Greg (et al.), Thanks for looking into it. And thanks to Paolo, who gave me a good workaround suggestion – which was to desymmetrise the spirocyclic centre by modifying the isotope on one of the neighbours. This is good for attended processing of single molecules, but not so good for

Re: [Rdkit-discuss] Stereochemistry issue for spirocycles/pseudochiral centres(?)

2017-02-08 Thread Greg Landrum
To further help with the work-around: it's safe to sanitize the molecule, but you cannot call Chem.AssignStereochemistry(), which the SMILES parser does when you tell it to sanitize. Here's an example from your gist: In [2]: m3 = Chem.MolFromSmiles('O[C@H]1CC[C@]11CC[C@@](Cl)(Br)CC1',