Re: [Rdkit-discuss] can't kekulize molecule

[Greg L. <gre...@gm...>]

The two primary functions that handle the current atom type semantics
are cleanUpMol2Substructures() and guessFormalCharges() here:
https://github.com/rdkit/rdkit/blob/master/Code/GraphMol/FileParsers/Mol2FileParser.cpp


On Wed, Aug 16, 2017 at 8:00 PM, Francois BERENGER <
ber...@bi...> wrote:

> On 08/16/2017 06:14 PM, Greg Landrum wrote:
>
>>
>> On Wed, Aug 16, 2017 at 3:55 AM, Francois BERENGER <
>> ber...@bi... <mailto:ber...@bi...>>
>> wrote:
>>
>>     On 08/16/2017 03:36 PM, Greg Landrum wrote:
>>
>>         The RDKit Mol2 parser is really only validated for the atom
>>         types generated by corina. I'm not surprised that the ouput from
>>         open babel would not be understood. This is documented:
>>         http://rdkit.org/docs/api/rdkit.Chem.rdmolfiles-module.html#
>> MolFromMol2File
>>         <http://rdkit.org/docs/api/rdkit.Chem.rdmolfiles-module.html
>> #MolFromMol2File>
>>
>>     It would be really nice if open babel MOL2 output could directly be
>> read
>>     in by rdkit.
>>
>> Adding this support is not an impossible task for someone who understands
>> the open babel interpretation of the Mol2 atom types. Nik's code for
>> dealing with the cleanup of the corina atom types is quite well documented
>> and creating a bunch of test cases using OpenBabel would be pretty
>> straightforward. It would take time and care though.
>>
>
> Can you point out that code?
>
> I may have a look one day.
>
>
> I'd guess that in the end it's easier and more straightforward to just let
>> open babel do the translation.
>>
>>     I often find myself running
>>     $ obabel in.mol2 -O out.sdf
>>     just for that purpose.
>>
>>
>> The question I always end up asking here is: Why do you have open babel
>> mol2 files in the first place?
>> If you're reading those into another piece of software (the usual
>> answer): are you sure that the other software and open babel interpret the
>> atom types the same way? Really sure?
>>
>> -greg
>>
>>