Dear Christoph,

I agree comma separated InChI-s are not good idea and will fix this ASAP.

Best regards,
Nina

On 22 March 2011 19:49, Christoph Helma <helma at in-silico.ch> wrote:

> Dear Nina, Vedrin,
>
> I have started to write ToxCreate tests to integrate Ambit datasets and
> ran across an issue with the Compound representation in Dataset
> http://apps.ideaconsult.net:8080/ambit2/dataset/9 (might apply to other
> datasets as well, but I did not check).
>
> If I ask for the InChI representation, I receive a comma separated list
> of InChIs and not a single one. I would expect however a single InChI
> per compound/conformer, because they are intended as unique identifiers
> for chemical structures (and I need them for that purpose in our
> services/libraries).
>
> Looking at examples (e.g.
> http://apps.ideaconsult.net:8080/ambit2/compound/10286/conformer/410062)
> suggests that you are providing separate InChIs for each InChI layer. I
> would suggest to provide the most specific layer that is available for a
> particular compound/conformer and leave it to the user to strip
> unnecessary information.
>
> What do you think?
>
> Best regards,
> Christoph
> _______________________________________________
> Development mailing list
> Development at opentox.org
> http://www.opentox.org/mailman/listinfo/development
>