Re: [Open Babel] Convert Molecule Structure to IUPAC

[qiancheng s. <qia...@gm...>]

Dear Geoffrey,
         I appreciate your help very much ! Thank you for your time.
         At first I thought it is a very simple problem, so I am confident
to find such a tool when a chemist friend asked me for help. But maybe you
are right, there is no such opensource tool exists.
         I have tried many tools, and maybe I should give a list and share
my experience on this problem.

         1. OPSIN , an open parser for IPUAC name. It can convert IUPAC name
to structure but not vice versa.
             (P.S. You can also use OPSIN with CDK to generate molecule
images from IUPAC name.)
         2. ChemDraw. It can generate IUPAC name from structure, but batch
mode seems to be inavailable.(not opensource)
         3. Openbabel.  It can convert structure to InChI name, but not
IUPAC name.
         4. OpenEye. Yes, it has a great tool to do this task, but not
opensource.
         5. Marvin Sketch. You can insert such IUPAC Name while drawing a
molecule, but not opensource.

Thank you for all your help!!


Shen

2011/4/5 Geoffrey Hutchison <ge...@ge...>

>
> On Apr 4, 2011, at 4:31 AM, qiancheng shen wrote:
>
> > Thank you very much!! But all these tools are not suitable for
> me......Any other ideas?
>
> You also asked on CCL.net and several people also pointed you to the NIH
> Chemical Resolver. Noel's link pointed out that Cinfony and Webel can
> provide you with a Python solution to query the Resolver and get a name. It
> won't work for all molecules, only the 16 million or so in PubChem.
>
> Put frankly, I dont expect this be solved in the open source space in the
> near term, unless the NIH or other government organization sponsors it. It's
> tedious, difficult work to code an IUPAC naming system. It has lots of
> tricky corner cases (even the commercial ones have problems). People are
> willing to pay money for such a solution, so there's a clear capitalist
> incentive to write something like Lexichem, sell it, and be financially
> rewarded for the work.
>
> Name to structure has a clear benefit for OPSIN and is technically an
> easier problem: parse the chemical name into determining structure. There
> may be many names for one structure, but that's not a big deal. OPSIN was
> also used to solve an obvious problem -- machine parsing journal articles
> for data mining.
>
> Structure to name has been requested for Open Babel, basically forever, but
> IMHO it's out of the scope of the project.
>
> Cheers,
> -Geoff