Re: [Rdkit-discuss] Inchi/smiles conversion issue

[Jennifer H. <jen...@gm...>]

Dear Alexis,

if you calculate the Standard Inchi it is invariant to tautomers (see 
here: https://www.inchi-trust.org/technical-faq-2/#6.1). Therefore the 
information which tautomer was converted is lost due to the Inchi 
conversion. If you want to keep the tautomer information you need to use 
the fixedH attribute for the inchi. But beware this makes it a non 
standard Inchi, and thus might not be comparable to other Inchis.

Hope this helps,

Jennifer

On 18.06.19 12:59, Alexis Parenty wrote:
>
> Dear RdKiters,
>
> Why is it that the stable tautomer of the following structure is lost 
> during inchi/smiles conversion?
>
> image.png
>
> mol = Chem.MolFromSmiles(*"Cc1ccc([nH]nc2)c2c1"*)
> inchi = Chem.MolToInchi(mol)
> mol = Chem.MolFromInchi(inchi)
> smiles = Chem.MolToSmiles(mol)
> print(smiles)
>
> /==> Cc1ccc2n[nH]cc2c1/
>
> The H has shifted on the wrong Nitrogen…
>
> Interestingly, if you remove the methyl, the shift no longer happens:
>
> mol = Chem.MolFromSmiles(*"c1([nH]nc2)c2cccc1"*) inchi = Chem.MolToInchi(mol) mol = Chem.MolFromInchi(inchi) smiles = 
> Chem.MolToSmiles(mol) print(smiles) ==> *c1([nH]nc2)c2cccc1*
>
> Same issue for any secondary amides: if you pass the smiles of a 
> secondary amide, you end-up with the following unstable tautomer:
>
> image.png
>
> Thanks,
>
> Alexis
>
>
>
> _______________________________________________
> Rdkit-discuss mailing list
> Rdk...@li...
> https://lists.sourceforge.net/lists/listinfo/rdkit-discuss