See also: IRC log
Michel: Agenda 1) demo, like Iker, then 2) paper
Iker, can you dial in?
<iker> one moment please
<iker> i am having problems with my VoiP service
Michel: Discussion of paper
paper from Chest http://chestjournal.chestpubs.org/content/133/6_suppl/160S.full.pdf
Bosse: Use case in a care
situation might be difficult
... could develop how info could help in pharma
... interested in making case for pharma; even if just around clinical trials
<michel> action item - Bosse to develop pharma-oriented use case for pharmacogenomics - will be nice to tie to a patient use case
Elgar: Can look at algorithms
that have already been developed
... everything ties into the use case
Michel: Portion of use case: doctor gathers info, warfarin-dosing.org(?)
Elgar: Interested in analysis
... want to understand where others did not have sufficient info
... have data, mash them up, then drill down; if we use the data, can we deal w. one of the algorithms than they did?
Matthias: Editors sent
requirement, top of google doc
... review or research article?
Scott: Will review original email.
<BobF> Michel: http://www.warfarindosing.org is the correct URL
Scott: suggested that we do pgx paper
<mscottm> The answer to this paste was yes: >>>>> Thank you for the invitation to write on the �Development of Multi-SNP
<mscottm> >>>>> Predictive Models in Pharmacogenomics�. We have been working on
<mscottm> >>>>> incorporating SNP's into our Translational Medicine Knowledge Base in
<mscottm> >>>>> one of the task forces of the W3C Health Care and Life Sciences
<mscottm> >>>>> Interest Group (HCLS IG). We would like to write a paper about an
<mscottm> >>>>> approach for integrating SNP's into multidisclipinary studies for
<mscottm> >>>>> translational medicine. Would that fit your needs?
ericP and Michel in discussion about false negatives, false positives
ericP: Sounds like low yield use
... machine gets to make decision and it always errs in one direction
... radiologist did not match microcalcification
... so machine can help where you otherwise might have missed
Elgar: This is clinical decision
support, part of what we are trying to represent
... based on this genotype, we expect that patient is higher- or lower-metabolizer
... original idea is to focus on clinical decision support alone
... research, etc, other scenarios can use one pgx representation
... patient can be influenced by warfarin-genotype
Scott: Started on breast cancer:
reason to go to warfarin is that bc is only expression
... expression profiles are more challenging to model than SNPs
... warfarin use case is more clear cut
... agree w. Matthias; use of pgx for translational medicine purposes
... also come up w representations for other biomarkers?
ericP: If is has both false negs and false positives, clincian always has to monitor patient anyway
Scott: Tumor stratification has a nicer (skit?)
ericP: Vigilance is use case here?
Michel: Dosage ranges do change; standard warfarin dosing is wrong > 50%
Matthias: Modeling all this in OWL is nontrivial
Michel: We can highlight this in
paper: chance of hurting patient w/o info
... warfarin on pharmGKB site, need to convert to RDF
BobF: On a call w pharmGKG, talking about representing in RDF; want to monitor this
ericP: converting vs chartering :-)
<iker> sorry, but i am definitely unable to join the conversation
Michel: can you do sparql
... share a raw RDF file.
Web Protege is a good idea.
Michel: Would need to understand axioms w. web Protege
Scott: Need your own server.
Michel: Protege 3 has client-server, configure acct etc
<iker> we have got an instance of webprotege with breast cancer ontology at http://188.8.131.52:8080/webprotege/WebProtege.html
Scott: Would like to get feedback on web Protege
<iker> you can check
Iker: Could you set up a project for us?
<michel> iker would you be able to setup an instance of webprotege for our project?
<iker> yes, no problem at all
Matthias: Easier to use a good
client than just a doc
... Web Protege has notes etc
Scott: PatientsLikeMe has written back. "What are your specific areas of interest?"
<iker> Would we start the project from scratch or do we have any ontology already implemented?
<Adrien> on google doc there is a diagram that will help start the ontology
This is scribe.perl Revision: 1.136 of Date: 2011/05/12 12:01:43 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) No ScribeNick specified. Guessing ScribeNick: Bob Inferring Scribes: Bob WARNING: No "Topic:" lines found. Default Present: Bob_Powers, mscottm, [IPcaller], Tony, +184.108.40.206.aaaa, +1.781.431.aabb, ericP Present: Bob_Powers mscottm [IPcaller] Tony +220.127.116.11.aaaa +1.781.431.aabb ericP WARNING: No meeting title found! You should specify the meeting title like this: <dbooth> Meeting: Weekly Baking Club Meeting WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth Got date from IRC log name: 19 May 2011 Guessing minutes URL: http://www.w3.org/2011/05/19-hcls2-minutes.html People with action items: WARNING: No "Topic: ..." lines found! Resulting HTML may have an empty (invalid) <ol>...</ol>. Explanation: "Topic: ..." lines are used to indicate the start of new discussion topics or agenda items, such as: <dbooth> Topic: Review of Amy's report[End of scribe.perl diagnostic output]