HCLSIG/PharmaOntology/Meetings/2009-10-15 Conference Call
Conference Details
- Date of Call: Thursday October 15 2009
- Time of Call: 12:00pm - 1pm ET
- 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: 42572 ("HCLS2").
- IRC Channel: irc.w3.org port 6665 channel #HCLS2 (see W3C IRC page for details, or see Web IRC)
- Mibbit instructions: go to http://www.mibbit.com/chat and click the server link. Enter irc.w3.org:6665 into that box, enter a nickname, and enter #HCLS2 for the channel
- Duration: 1h
- Convener: Susie
Agenda
- Ontology (namespace) - Colin
- Use case - Chris
- Data sources - Anja
- User interface - Susie
- Paper - Susie
- AOB
Minutes
In attendance: Lynn, Christi, Matthias, Jim McGurk, Michel, Joanne, Elgar, Susie, Colin, Chris, Trish, Jun, Bosse
susie: what progress on ontology?
colin: telecon between colin, elgar, michel & joanne
colin: minutes: http://docs.google.com/Doc?id=dc9qc59t_17f244jtdv&btr=EmailImport
elgar: minutes from Mon 10/12 meeting on TMO: http://esw.w3.org/topic/HCLSIG/PharmaOntology/Meetings
colin: protege 4, SVN on google code project, external terms in tmo-external.owl, whereas tmo.owl is just TMO terms
colin: comments about unplaced items were solicited
colin: try to get more xrefs to other ontologies
colin: ontological coercion -> use terms that are in thesauri like mesh, but place them in such a way that we can reason about them
susie: things broken were fixed?
colin: yes
susie: w3c namespace will be available on monday by eric
susie: what else to be done?
colin: definitions for undefined terms
elgar: data, datasets, influence on final version of tmo?
colin: undefined terms need to be concretized in terms of what they mean based on the data that we plan to annotated with them
colin: biomarker has several definitions... which is the most appropriate depends on the data
colin: disambiguation
susie: we need to talk about data sources
susie: michel and I have come up with list
susie: pharmagkb (personalization), drugbank (drugs), sider (side effects), clinicaltrials.gov (clinical trials), diseasome (disease--gene correlations), medicare (what drugs are available on particular plans)
susie: also diagnostic information and electronic health record information
susie: Linked Data available for first 5.
susie: Anja Jentzsch is RDFifying the medicare set
susie: statements like is_available_on(aricept, medicare) - available by end of week
susie: ehr - assume this exists - open system called indivo developed at Harvard - schemata publicly available and used by google & microsoft
susie: asked them for fake patients but they weren't forthcomingso we will make up our own
bosse: was there personal health information in the COI task?
susie: ours focuses on AD; theirs on diabetes
susie: will contact them
susie: only 5 patients needed
joanne: Peter Tonellato at the Library for Personalized Medicine (http://lpm.hms.harvard.edu/), part of the Center for Biomedical Informatics @ Harvard creates fake patients, "avatars" - I can look into this
susie: we do need cognitive information, genetic information e.g. ApoE, SNPs implicated in recent Nat. Genet. paper posted on wiki
chris: what about toxicity/efficacy?
chris: and SNPs?
susie: JnJ and ?Illan jointly own a drug and trials of this have shown differing effects for people with ApoE
susie: also important to make up ADME data, variant version of cytochrome P450
susie: another option - Lilly developed DMet chip with key genes
susie: I'll make up data and pass to joanne and ping coi people
joanne: inquire about suggestions Peter may have about generating fake patient data
susie: chris has pointed me to data sources with diagnostic information, also academic papers
susie: we could run unstructured text through bioportal
susie: and then triplify that
susie: and make sure we're using those terms in the fake patient data
michel: key Alz / drug / variant / outcome papers:
michel: http://www.pharmgkb.org/do/serve?objId=569429393&objCls=RelationshipReport#publications
michel: http://www.pharmgkb.org/do/serve?objId=569436230&objCls=RelationshipReport#publications
michel: http://www.pharmgkb.org/do/serve?objId=569434972&objCls=RelationshipReport#publications
michel: http://www.pharmgkb.org/do/serve?objId=569429027&objCls=RelationshipReport#publications
chris: what about real medical sources?
susie: much harder than in bioinformatics; we need to be credible rather than working with really really difficult real sources
susie: PharmaSurveyor (a company) run text-mining on top of semantic web data looking for drug-drug interactions
susie: and michel has kindly volunteered to stick data in a triplestore
susie: and then map to ontology
michel: the lodd triplestore : http://bio2rdf.semanticscience.org:8017/sparql/?default-graph-uri=&should-sponge=&query=select+distinct+%3FConcept+where+{[]+a+%3FConcept}&format=text/html&debug=on&timeout=
susie: and then run application from that
michel: pharmgkb - http://bio2rdf.semanticscience.org:8016/sparql/
susie: Trish - do you have time to run data sources through Bioportal by say Tuesday?
susie: also by then Anja's medicare data should beready
trish: sure
susie: any questions?
susie: and now, the interface
susie: option (1) pubby
susie: option (2) Anja's top questions interface
susie: option (3) Jun's AD application
jun: doable but really difficult by end of month
susie: option (4) bosse's experience of OntoText interface on LarKC
bosse: not in time for paper
joanne: also will be travelling last 2 weeks in October.
susie: ok so Anja could provide top questions, or we could use Pubby as extension unlikely
joanne: need not be polished
susie: need to think about what the interface would look like
susie: jun - could you work with us on look of interface?
jun: sure!
jun: available till thursday
bosse, elgar, michel: all interested
susie: let's schedule for lodd slot 1100 edt weds
susie: october 21st
susie: last item is paper itself susie: google doc available - colin what's the position?
colin: written aspects of the introduction, the aim, scope, general outline
colin: ontology, design methology, mappings
colin: moved lots of chris's stuff to new context section
colin: data sources & their descriptions
susie: chris and susie will work on conclusion
susie: when do we ask for an extension?
joanne: no harm in asking upfront
susie: we're getting there