Re: [Rdkit-discuss] sanitization removes Hs - is this expected?

[Michal K. <mic...@gm...>]

Thanks Greg, this is exactly what I wanted to know. Would you consider
adding an optional removeHs argument to MolFromSmiles(), as in
mol/mol2/sdf parsers?
Best wishes,
Michal

On 25 February 2014 04:23, Greg Landrum <gre...@gm...> wrote:
> Hi Michal,
>
> On Mon, Feb 24, 2014 at 4:48 PM, Michal Krompiec <mic...@gm...>
> wrote:
>>
>> Hello, I have just noticed this:
>> >>> Chem.MolToSmiles(Chem.MolFromSmiles("[H]c1c([H])sc([H])c1[H]"))
>> 'c1ccsc1'
>> >>>
>> >>> Chem.MolToSmiles(Chem.MolFromSmiles("[H]c1c([H])sc([H])c1[H]",sanitize=False))
>> '[H]c1sc([H])c([H])c1[H]'
>> >>>
>> >>> Chem.MolToSmiles(Chem.RemoveHs(Chem.MolFromSmiles("[H]c1c([H])sc([H])c1[H]",sanitize=False)))
>> 'c1ccsc1'
>> >>> Chem.MolToSmiles(Chem.MolFromSmiles("[H]c1cscc1[H]"))
>> 'c1ccsc1'
>> >>> Chem.MolToSmiles(Chem.MolFromSmiles("[H]c1cscc1[H]",sanitize=False))
>> '[H]c1cscc1[H]'
>>
>> Is it the expected behaviour? Why does sanitization remove hydrogens?
>>
>> Is it controlled by any of the SanitizeFlags?
>
>
>  It is the expected behavior. When sanitization is turned on, the SMILES
> parser actually calls "RemoveHs"; this removes the hydrogens from the graph
> and then sanitizes the molecule.
>
> If you do not want the Hs removed, you can tell MolFromSmiles to skip the
> sanitization (which also skips the RemoveHs) and then sanitize yourself::
>
> In [3]: m=Chem.MolFromSmiles("[H]c1c([H])sc([H])c1[H]",sanitize=False)
>
> In [4]: Chem.SanitizeMol(m)
> Out[4]: rdkit.Chem.rdmolops.SanitizeFlags.SANITIZE_NONE
>
> In [5]: print Chem.MolToSmiles(m)
> [H]c1sc([H])c([H])c1[H]
>
> I hope this helps,
> -greg
>