Forums

Nicolas,

>> If you have connectivity to the sboTerm and, (as some suggest) it has
to
>> have a mathematical translation attached to it so that the symbols are
in
>> a standard order, then there is no excuse for not writing the mathML.
>> If you don't have the connectivity, a <notes> <B>kineticlaw is mass
>> action</B></notes> avoids problems of spelling errors that we have
>> sometimes seen when using some of the named equations. It is at least
as
>> informative as the kineticLaws with only sboTerms.

>Certainly not. If that was the case, nobody would have created Gene
>Ontology, and that would not have become in a few years one of the most
>successful achievements of bioinformatics. A notes element with such a
>sentence has hardly any semantic content. First of all, no software but
>the most advanced text-mining tool can extract the fact that it is
dealing
>with mass-action (becasue of spelling, encoding, foreign languages etc.).
>Second, a biologist tend to use an unconstrained language. In
biochemistry
>this is even more tricky. 99.999 % of the articles speak of
>"Michaelis-Menten" while they mean "Briggs-Haldane". It is obvious that
>the modellers would do the same in the unconstrained annotation.

Without the connectivity, you're not only going to get
"Michaelis-Menten", you're going to get "Michaelis-Menton",
"Michelis-Menten", etc. as we have in the past. There is no checking or
constraining from external sources, so GO is totally irrelevant. Kindly
explain why this is better than a notes annotation when BOTH are
unconstrained.

If you do connect to the site to retrieve the sboTerm, I repeat
that you do not have the excuse not to write the mathML as well.



Hugh Spence
GSK Scientific Computing and Mathematical Modelling
Medicines Research Centre
Gunnels Wood Road
Stevenage
UK
SG1 2NY






"Nicolas Le Novère" <lenov@ebi.ac.uk>
Sent by: sbml-discuss-bounces@caltech.edu
05-Jan-2006 19:02
Please respond to "SBML Discussion List" <sbml-discuss@caltech.edu>


To
"SBML Discussion List" <sbml-discuss@caltech.edu>
cc

Subject
Re: [sbml-discuss] SBML L2v2 specification vote #4: References to
controlled vocabularies







> If you have connectivity to the sboTerm and, (as some suggest) it has to
> have a mathematical translation attached to it so that the symbols are
in
> a standard order, then there is no excuse for not writing the mathML.
> If you don't have the connectivity, a <notes> <B>kineticlaw is mass
> action</B></notes> avoids problems of spelling errors that we have
> sometimes seen when using some of the named equations. It is at least as
> informative as the kineticLaws with only sboTerms.

Certainly not. If that was the case, nobody would have created Gene
Ontology, and that would not have become in a few years one of the most
successful achievements of bioinformatics. A notes element with such a
sentence has hardly any semantic content. First of all, no software but
the most advanced text-mining tool can extract the fact that it is dealing
with mass-action (becasue of spelling, encoding, foreign languages etc.).
Second, a biologist tend to use an unconstrained language. In biochemistry
this is even more tricky. 99.999 % of the articles speak of
"Michaelis-Menten"
while they mean "Briggs-Haldane". It is obvious that the modellers would
do the same in the unconstrained annotation.


--
Nicolas LE NOVÈRE, Computational Neurobiology,
EMBL-EBI, Wellcome-Trust Genome Campus, Hinxton, Cambridge, CB10 1SD, UK
Tel: +44(0)1223 494 521, Fax: +44(0)1223 494 468, Mob: +33(0)689218676
http://www.ebi.ac.uk/~lenov