HCLSIG/LODD/Meetings/2010-02-03 Conference Call

Conference Details

Date of Call: Wednesday February 3, 2010
Time of Call: 11:00am Eastern Daylight Time (EDT), 16:00 British Summer Time (BST), 17:00 Central European Time (CET)
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: ~1h
Convener: Susie

Agenda

Data updates - Anja, Jun, Matthias, Egon
Mapping to drugs/intervention - Oktie
Mapping experimental data - All
Outreach (TCM, Bio-Ontologies SIG, ACS, AMIA) - All
AOB

Minutes

Attendees: Matthias, Kei, Oktie, Joanne, Egon, Jun

Data Updates
<OktieH> http://linkeddata.cs.toronto.edu/linkedfda
<Susie> Oktie: Linked FDA data is already ready
<Susie> Oktie: Can talk about it more during the next call
<Susie> Susie: Having your colleague present sounds good
<Susie> Egon: Egon did a schema walk through on ChEMBL
<Susie> Egon: Looking at protein classification scheme
<Susie> Egon: Struggling to make it available in a useful way
<Susie> Egon: Looking at indexing schemes
<Susie> Egon: Lots of data
<Susie> Egon: Activities is over 1GB in RDF/XML, over 2GB in total
<egonw> Egon: chembl sparql end point: http://pele.farmbio.uu.se/chembl/sparql/
<jun> This might be helpful for you:

http://www.openlinksw.com/dataspace/dav/wiki/Main/VirtBulkRDFLoaderExampleMultiple

<jun> Egon: I have loaded more than 1GB RDF triples to Virtuoso
<Susie> Egon: Interesting as links proteins to assays
<egonw> Egon: ChEMBL links: protein, assay, activity, molecule
<egonw> Egon: and all to DOIs
<Susie> Matthias: Matthias has written to data providers
<Susie> Matthias: But they didn't reply and provide access to data
<Susie> Matthias: So rather short of data relating to chinese data
<Susie> Matthias: Nice data sources exist but are not available to integrate
<Susie> Matthias: Focus paper on structured data, mined data and TCM
<egonw> jun: I have been looking at this page for the indexing: http://docs.openlinksw.com/virtuoso/rdfperformancetuning.html#rdfperfgeneraldbpedia
<Susie> Matthias: Although there isn't new data to add to the collection, thinking about how to combine in the best
<Susie> Matthias: Very interested in how best to present the data
<Susie> Kei: wonder if papers provide supplementary dat asets that you can use
<Susie> Matthias: Old school, so no supplementary data
<Susie> Kei: Wikipedia/DBpedia has broad classification of herbs that could be useful
<Susie> Kei: Another source that could be useful for inferring could be the NCI taxonomy
<Susie> Matthias: Plan to focus on small set of plant species that appear promising
<Susie> Matthias: Present aggregated way so that the researcher can look at the data and make inferences in their mind
<Susie> Matthias: Goal is for researchers to get insights more quickly by looking at the aggregated paper rather than by reading the papers
<matthias_samwald> http://less.aksw.org/browse
<Susie> Matthias: Looked at this templating language, where looking at simple views from linked data or results from sparql query
<Susie> Matthias: Looks nice but might be too limited for our purposes
<matthias_samwald> http://sig.ma/
<Susie> Matthias: Also looked at Sigma
<Susie> Matthias: It allows nicely formatted mashup of data for a certain query
<egonw> matthias_samwald: looks interesting...
<matthias_samwald> http://samwald.info/antidepressant-effects-of-magnolia-officinalis-aggregated-linked-data/
<egonw> matthias_samwald: is it free, or open source?
<Susie> Matthias: Contains data from Jun's TCM data set and DBpedia
<Susie> Matthias: Limited for our use case
<Susie> Matthias: As want to view data that spans many triples at once
<Susie> Matthias: While Sigma more focused on showing a particular triple
<Susie> Matthias: So not always helpful
Mapping Experimental Data
<Susie> Susie: Approach for understanding best approaches for mapping linked data
<Susie> Susie: Create a doc with questions
<Susie> Susie: And then begin to fill in with solutions
<Susie> Susie: Can then discuss what would make best practices
<jluciano> Joanne: who decides what the answer is? Is there more than one answer (for different interperations of the data)?
<Susie> Susie: ADNI is an example of such data
<Susie> Susie Contains information about patients, cognitive scores, brain scans, CSF data, etc.
<Susie> Susie: What are best practices for mapping data about a patient, on a particular visit, having an MRI, that is of 1.5T, where hippocampal volume is measured, and a number derived using a particular image analysis approach
Outreach
<Susie> Matthias: TCM Paper
<Susie> Matthias: Deadline extended to mid-Feb
<Susie> Matthias: Going to Japan this w/e for the biohackathon
<Susie> Matthias: So time becomes limited
<egonw> Egon: cool :) heard good stories about that meeting
<Susie> Matthias: Aiming to get it finished this w/e
<Susie> BioOntologies SIG
<Susie> http://sites.google.com/site/bioontologies/
<Susie> Susie: Deadline is April 16.
<Susie> Egon: CFP for the linked data workshop at ACS went out
<egonw> http://chem-bla-ics.blogspot.com/2010/01/semantic-chemistry-with-resource.html
<Susie> Egon: March 28 deadline for submissions
<Susie> Egon: May be half or full day depending on numbers of submissions
<Susie> Egon: Meeting is in August in Boston
<Susie> Egon: Looking for anything related to chemistry
<Susie> Egon: ACS covers biochemistry under chemistry too
<Susie> Egon: Have a student working on substructure mining
<Susie> Egon: Just ChEMBL isn't enough for a paper
<Susie> Egon: Just need to submit an abstract for consideration for a talk
<Susie> Talks may be written up as papers for publication in a journal
<Susie> Protein structures or small molecules will need to be a focus
<egonw> Egon: let me stress, I'd love to see a HCLS submission... but being organized, someone else need to bootstrap that
<egonw> Egon: umm... but being organizeR..
<egonw> Egon: for ACS :)