W3C home > Mailing lists > Public > public-semweb-lifesci@w3.org > May 2009

Re: drug side effects

From: Matthias Samwald <samwald@gmx.at>
Date: Fri, 29 May 2009 11:12:37 +0200
Message-ID: <2AB19D45D36B48A2AC438C4FC8ABE65C@ms>
To: "public-semweb-lifesci" <public-semweb-lifesci@w3.org>
Cc: "Holger Stenzhorn" <holger.stenzhorn@deri.org>
The conversion of SIDER to aTags is now available, see
http://esw.w3.org/topic/HCLSIG_BioRDF_Subgroup/aTags/datasets

A small excerpt of the aTags:
http://hcls.deri.org/atag/data/SIDER_atags_excerpt.html

The aTags re-use URIs from DBpedia, OBO disease ontology and OBO symptom 
ontology. Most statements link a DBpedia URI (a drug) to an OBO URI (a 
disease/symptom).

As a simple example, the following query shows statements in SIDER that deal 
with some kind of CNS disease:

SELECT ?atag_content ?some_cns_disease
WHERE {
  ?atag sioc:content ?atag_content .
  ?atag sioc:topic ?some_cns_disease .
  ?some_cns_disease rdfs:subClassOf disease_ontology:DOID_331 .
  }
LIMIT 500

Results: http://tinyurl.com/SIDER-about-CNS

Some of the results look funny, for example, Memantine has 'Alzheimer's' 
listed as a side-effect. However, Memantine is actually used as a treatment 
for Alzheimer's. I have to look into that -- maybe SIDER should just be seen 
as providing associations between drugs and diseases/symptoms, without 
specifying whether they cause or cure the disease? This is also the case 
with the TCM datasets that Jun converted to RDF recently.

Note that the SPARQL endpoint currently only contains a part of the data, 
possibly the Virtuoso Sponger has problems with parsing large RDFa files. 
Also note that others also have recently converted SIDER into RDF as well 
(Anja Jentsch, Peter Ansell).

Cheers,
Matthias Samwald

DERI Galway, Ireland
http://deri.ie/

Konrad Lorenz Institute for Evolution & Cognition Research, Austria
http://kli.ac.at/ 
Received on Friday, 29 May 2009 09:24:47 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 18:00:55 GMT