HCLSIG BioRDF Subgroup/Meetings/2009-07-06 Conference Call

From W3C Wiki

Conference Details

  • Date of Call: Monday July 6, 2009
  • Time of Call: 11:00 am Eastern Time
  • Dial-In #: +1.617.761.6200 (Cambridge, MA)
  • Dial-In #: +33.4.89.06.34.99 (Nice, France)
  • Dial-In #: +44.117.370.6152 (Bristol, UK)
  • Participant Access Code: 4257 ("HCLS")
  • IRC Channel: irc.w3.org port 6665 channel #hcls (see W3C IRC page for details, or see Web IRC)
  • Duration: ~1 hour
  • Frequency: bi-weekly
  • Convener: Kei Cheung
  • Scribe: Rob Frost

Attendees

Rob Frost, Kei Cheung, M. Scott Marshall, Lena Deus, Satya Sahoo, Jun Zhao

Regrets

Matthias Samwald

Agenda

  • TCM (Traditional Chinese Medicine) update
  • Query Federation -- Microarray data and its integration with other data sets

Minutes

<rfrost> First agenda item, TCM update

<rfrost> Jun taking lead on project

<rfrost> collaboration between BioRDF and LODD

<rfrost> Jun: poster is in good shape, decided which portions each person will fill in

<rfrost> Kei: exploring linked data approach, how it can be used to connect western drug data and chinese herbal data

<rfrost> Kei: link using string matching on common gene names

<rfrost> Kei: various issues on linking genes by gene names between various datasets

<rfrost> Kei: can also link by drug names

<rfrost> kei: interesting outcome is estabilishing best practices regarding this type of dataset integration

<rfrost> Kei: integration of herbal medicine is complementary therapeutic approach to western medicine

<rfrost> Scott: linking by gene name is difficult due to large number of synonyms

<rfrost> mscottm: when you link by drug names do you then retrieve compounds?

<rfrost> kei: yes

<rfrost> mscottm: chemspider, potentially relevant dataset

<rfrost> mscottm: may be useful for recognizing drug compounds referenced in literature

<rfrost> mscottm: drug name->compound and then identify the compound name via chemspider

<mscottm> http://aida.science.uva.nl:9999/search/

<mscottm> http://www.csw.inf.fu-berlin.de:4039/sesame

<mscottm> Here's another: http://ws.adaptivedisclosure.org/search/

<rfrost> mscottm: can use AIDA to connect to Berlin endpoint - can see SenseLab dataset, TCM is not appearing

<rfrost> Jun: will forward SPARQL endpoints and other information to Scott to help further investigate issue

<rfrost> kei: discussed possible future directions for TCM during last call

<rfrost> kei: one option was to contact Huajun the possibility of integrating TCM datasets that are written mainly in Chinese

<rfrost> kei: did contact and he is interested in participating in joining the BioRDF call and giving presentation

<rfrost> Query federation topic

<rforst> kei: is there a publication date for BMC special issue

<rfrost> mscottm: does not know if a specific publication date has been set

<rfrost> kei: mentioned possible future directions for query federation in paper

<rfrost> kei: one potential area for investigation is the integration of new data sources

<kei> http://np2.ctrl.ucla.edu/np2/viewProject.do?action=viewProject&projectId=433773

<rfrost> kei: sent out email outlining microarray data

<rfrost> kei: link for microarray experiment "Project steph-affy-human-433773"

<mscottm> got it

<rfrost> kei: one of multiple neuroscience related microarray experiments; each experiments includes links to associated publications

<rfrost> kei: how to represent this data in RDF/OWL?

<rfrost> kei: which experimental data can be considered metadata/provenance?

<rfrost> mscottm: create RDF triples from HTML markup?

<rfrost> Project data included MAGE-ML

<rfrost> should investigate mapping from MAGE-ML to RDF

<rfrost> will require multiple existing ontologies to model associated data

<rfrost> kei: references to NIF ontologies

<rfrost> Jun: large overlap between MAGE ontology and NCI Thesaurus

<rfrost> kei: task: identify appropriate ontologies for modeling experimental information

<mscottm> http://bioportal.bioontology.org/

<mscottm> http://www.bioontology.org/wiki/index.php/Using_NCBO_Technology_In_Your_Project

<rfrost> rfrost: attempt to use Annotator on experiment HTML/publication/MAGE-ML data

<rfrost> Lena: has experimented with bioportal webservices

<kei> Rob, Satya, and Lena may start exploring the use of ncbo ontology services in the context of the microarray example

<rfrost> lena: RESTful services

<rfrost> kei: raw microarray data may not be appropriate for use on query federation task

<rfrost> kei: attempt to integrate gene expression data in these publications with ABA and GeneSat

<rfrost> lena: mage has concept of levels of data

<kei> differentially expressed genes should be focused on first

<rfrost> mscottm: microarray data not as reliable as other methods

<rfrost> lena: appears that gene list in paper is not complete

<rfrost> (gene list in fig 3 & 4 in paper referenced in Kei's email)

<kei> lena will contact the author for genelists

<rfrost> kei: there are other papers/experiments covering other brain regions/conditions

<kei> possible use of atags?

<rfrost> mscottm: does MAGE-ML contain sufficient data to enable integration with other datasets?

<mscottm> <OntologyEntry category="CellType" value="layer III neurons" description="">

<rfrost> jena: MAGE-ML does not cover level 2 or level 3

<mscottm> <OntologyEntry category="OrganismPartRegion" value="Entorhinal Cortex" description="">

<rfrost> need to further investigate exact structure/content of MAGE-ML and alignment with other datasets/ontologies of interest

<LenaDeus> http://mged.sourceforge.net/ontologies/index.php

<rfrost> mscottm: federation scenario: federated query across existing HCLS datasets, SWAN. etc.

<rfrost> mscottm: talk to Paola from scientific discourse group about potentially relevant experiences

<rfrost> kei: could be good opportunity for collaboration with other task forces

<mscottm> http://www.ebi.ac.uk/gxa/

<kei> scott: atlas of gene expression

<rfrost> mscottm: EBI just launched an atlas of gene expression; may be a relevant dataset