[Biojava-l] SAX, DOM, XPath and Flat files

[Matthew Pocock]

On Wednesday 12 December 2007, Mark Schreiber wrote:
> Not a bad suggestion. I wasn't aware of how many formats are now in
> OWL. It would give us a pretty rapid route to pathway and microarray
> object models as well.

I expect that the majority of bioinformatics data that is currently stored in 
RDBMS or other highly-structured forms will be published in either raw RDF, 
or in OWL as an alternative to the current access methods. How much of this 
is 'syntactic' and how much of it will require some reasoning services to 
make sense of remains to be seen though.

>
> Jena seems to be a nice package to build upon. Unfortunately due to
> the 'striped' rather than nested nature of ontologies we can forget
> the SAX vs DOM argument. The Jena OntModel is all going into memory.
> Looks like informatics is going to be the memory pig of the next
> decade.

We tend to use the OWL-API maintained at manchester university. Did I mention 
someone talking about jar dependency explosions? Maven is the solution here, 
my friends.

>
> - Mark

Matthew