too many evidence instances?

0 views

Skip to first unread message

Alan Ruttenberg

unread,

May 31, 2006, 11:25:19 PM5/31/06

to Ken I Fukuda, BioPAX Manchester

Hi Ken,

I'm looking a little closer at TGFB superfamily that you sent to the
manchester list a bit ago.
No comprehensive comments yet (though very interesting - I am also
looking at the inoh version in your viewer :)

However one thing I noticed was that there seem to be too many
evidence instances. I count 233. Since each is a pair of paper and
evidence code, and there are 6 papers and 4 evidence codes, I would
expect that there would be at most 24 distinct possible evidences. I
could imagine more

(sparql '(:select (?value) (:distinct t) (?thing !bp2:EVIDENCE-CODE ?
value)) :kb (kb inoh-tgfb) :flatten t)

-> (!bp2:AS !bp2:IC !bp2:IEX !bp2:TAS)

(length (sparql '(:select (?evidence) (:distinct t) (?evidence a !
bp2:evidence)) :kb (kb inoh-tgfb)))

-> 233

More as I find it.

Regards,
Alan

Ken I Fukuda

unread,

Jun 1, 2006, 1:32:02 PM6/1/06

to Alan Ruttenberg, BioPAX Manchester

Hi Alan,

Thanks for pointing out.
All nodes can have pairs of a paper and an evidence code.
The problem is the XSLT creats a separate instance for each node.
We will fix this so that an evidence instance will be shared
among INOH nodes.