On Tue, Mar 9, 2021 at 6:15 PM Jason Biggs <jasondbi...@gmail.com> wrote:
> > On Tue, Mar 9, 2021 at 10:49 AM Maciek Wójcikowski <mac...@wojcikowski.pl> > wrote: > >> Hi Jean-Marc, >> >> I know you can draw them, but both SMILES and RDKit internally use two >> bonds (up/down) directions to assign the bond stereo, which means that >> there are not >> > > Fortunately the RDKit also allows you to set bond stereo without using the > bond directions, by setting the bond stereo flag to stereocis or > stereotrans and then setting the stereoAtoms to say which atoms are cis or > trans. Using bond directions for double bond stereo has always seemed > rather archaic and confusing to me. > Though I agree with you completely, unfortunately the IUPAC recommendation (which the RDKit does not follow, but needs to support) is that you should indicate unknown bond stereochemistry using a wavy bond (which the RDKit encodes as bond direction). Here are two excerpt images from the recommendations ( https://www.degruyter.com/document/doi/10.1351/pac200678101897/html) [image: image.png][image: image.png] -greg
_______________________________________________ Rdkit-discuss mailing list Rdkit-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
