Re: [Obo-format] ***UNCHECKED*** Re: OBO to OWL (and back)

[Dilvan M. <di...@st...>]

On 9/22/06, st...@in... <st...@in...> wrote:
> >And I do
> > not think the ID of the DbXref should point to the resource directly
> > (in the cases where you can derive a URI name from a Dbxref) as they
> > are two diferrent things.
>
> ok, last thought-is there a case for the dbname being a URI/anyURI?
> (as there is this problem with mapping dbname:acc ->URI)

   I think you are going to have the same problem here. One can write
any string as a Dbxref and still be a valid Dbxref in OBO. Even the
colon isn't mandatory in the standard.

> should be fine, i'll take a look later.

   They are there, I tried to highlight them using colors, but you
cannot change the color of individual words. Please take a look.

> I'm aware that the OBO files are in a big mix of formats, i found that
> opening in OBOEdit and saving under the 'strict' option was a way to
> fix some problems.....running tests over these files is a lot of work!!!

   I created a JUnit test program that reads all ontologies in the obo
repository and test the round trip in all of them. Some aren't even
well formed OBO files (I discard them). But I tested all that the OBO
parser is capable of reading and use the OBO diff to determine if they
survive the round trip or not.

   Alan Rector wants to be CC on the emails about mapping OBO->OWL (he
already has the spreadsheet address) and Chris has been away, so I
will try to write an email pointing out the topics we didn't reach
agreement yet.

   Cheers,


-- 
Dilvan de Abreu Moreira, Ph.D.     di...@st...
SMI-Stanford Phone: 650-725 6236     http://java.icmc.usp.br
Warning: I use a spam filter, some emails sent to me CAN be lost!