Re: [Rdkit-discuss] adding fragment to existing molecule

Hi Per, I can think of 2 approaches to solve this. The 1st is to have fragments of molecules that have an explicit connection point, i.e. OH[*] and C[*], and use RDKit's functionality of combining fragments. The 2nd is to use define a reaction for this using SMIRKS or Reaction SMILES, i.e.

Re: [Rdkit-discuss] problem with AssignAtomChiralTagsFromStructure

I don't think I understand the question. The Conformer object is normally attached to/associated with a mol object. -greg On Sun, Aug 6, 2017 at 11:41 AM, Per Jr. Greisen wrote: > Thanks - worked perfectly if I was to do this conformers is there an easy > way to transform

Re: [Rdkit-discuss] using rdkit to read in chembl23 1.7 million compounds

Hi Brian, It's not that surprising. The RDKit is stricter about allowing unreasonable chemistry that the tool the ChEMBL group uses to produce SMILES or mol blocks. There are always some molecules that the RDKit just won't process. If you are concerned and see any in that group of failures that

Re: [Rdkit-discuss] . Re: using rdkit to read in chembl23 1.7 million compounds

That molecule's SMILES is correctly rendered by RDKit, or at least by the version of RDKit behind Slack: [image: Inline image 1] -P. On Mon, Aug 7, 2017 at 3:54 PM, Bennion, Brian wrote: > The carbocations are in small heterocyclic molecules. see CHEMBL3815233 > > Brian >

Re: [Rdkit-discuss] . Re: using rdkit to read in chembl23 1.7 million compounds

Hello Peter, Great, that just made me realize that I was not using my most recent conda environment version of RDkit. I reread the 2D sdf file with the latest rdkit version and now only 31 molecules are tossed out by the SDMolsupplier in RDKit. 51 compounds had errors when reading in the