HCLSIG/PharmaOntology/Meetings/2009-11-12 Conference Call
Conference Details
- Date of Call: Thursday November 12 2009
- Time of Call: 12:00 - 1:00pm 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 Update - Colin
- * questions: (1) names - spacy or underliny? (2) what issues have been raised by data loading? (3) what to do about "outcome", "treatment safety", etc?
- Diagnostic Data - Michel
- PCHR Data - Susie
- Data Loading and Mapping - Michel
- Paper (intro, intro box, UI, etc.) - All
- AOB
Minutes
Attendees: Chris, Jim, Colin, Bosse, Michel, Joshua, Matthias, EricM, EricP, Christi, Trish, Elgar, Susie
in attendance: Susie, Chris, Colin, Bosse, Jim, Matthias, Michel, Joshua Phillips, Eric Miller, Trish, Elgar
susie: agenda - Colin to do ontology, Michel to do diagnostic data, Susie to do pchr data, Michel to do data loading and mapping
colin: have updated with role relations, Elgar's questions
colin: spaces or underlines?
susie: happy whichever?
elgar: underlines preferred
elgar: inconsistency between versions of protege - looks nicer with underlines but is mostly cosmetic
trish: it's a problem if the human-readable name is being used as the rdf id
trish: what are we usinf for names
colin: both rdfs:label and iao:IAO_115
susie: we need to make a choice - my bias to underscores
susie: can you guys take this offline? ACTION
colin: to bear in mind - what issues have been raised by data loading, if any?
colin: what to do about unassigned terms like "health outcome"?
susie: happy for you to pursue business about lifestyle
susie: would like to capture "health outcome"
susie: how did they respond?
colin: does the patient data have outcomes in it?
susie: in the multiple annual medicals
colin: so the outcome proceeds from the difference between successive encounters
susie: also correlation between snps/lifestyle and encounters
trish: is the concern that there are orphan classes?
colin: wondering what people think
susie: since outcome and safety are two of the most important things
elgar: we should make up our minds, given we have bfo
trish: tentatively place w/ curation status
all: agreed
colin: susie - do we have data that maps on to safety
susie: i'd like to add adverse events - could map on to safety
chris: i'll think about that
susie: we have SIDER
susie: are the patient records realistic?
colin: safety difficult as absence of bad things happening
elgar: general understanding is negative effects
chris: worth further discussion
bosse: not an adverse drug effect because may not be caused by a drug
susie: colin - are happy to get together w/ bosse, elgar & chris about safety
colin: yes
susie: DIAGNOSTIC DATA
susie: ------------------
michel: have taken text from panel 2, passed it through NCBO Annotator
michel: http://tm.semanticscience.org/describe/?url=http://bio2rdf.org/tmo:panel2/ffcdde6da9ecaa4a
michel: rdf'ified results
michel: link above tells us about "amygdala"
michel: linkage to ontology nebulous because lack information about meaning and where it came from
michel: once i understand how that works can load concepts into endpoint
susie: trish - does this supersede what you've done?
trish: did something similar but was a more specific subset. have been in touch w/ michel and maryann martone
elgar: trish - i put two random pubmed abstracts through annotator and was disappointed; did i use it correctly? is there documentation on entity recognition?
trish: just looks for preferred name + synonyms
elgar: do i have to select ontologies explicitly?
trish: there is a 'select all' option
susie: seems we're on track
susie: FAKE PATIENT DATA
susie
susie: Chris has put a lot of work into the different patients
susie: i have been updating Patient One (the template) - want to look at it one more time before publishing
susie: bosse - you've signed up for patient 3
bosse: much harder than i expected - where do i get drug code
chris: just NDC
susie: I was following Peter Kos - also asked him re genomic information but he had to go to 3 web sites
susie: keep as simple as possible
susie: bosse - keep plugging away
susie: Peter Kos has volunteered to fill in some of the gaps - hopefully by splitting load we can get it done quicker
chris: genomic section difficult
susie: some of the genotyping can be repeated across patients
christi: I got the lab results in
christi: doing Downs Syndrome example
susie:christi has made great progress, anja has flu, matthias?
matthias: very complicated - bit puzzled about how xml should be structured
chris: how do we compare?
susie: getting uploaded to wiki
susie: jim - what progress?
jim: have been working but lack access to wiki
susie: jim- could you do patient 4?
jim: sure
susie: definitely making progress though more of a task than I'd expected. but very worthwhile
jim: happy to do xslt
susie: look at patient one
chris: what one do i start on?
susie: christi's
susie: DATA LOADING AND MAPPING
michel: just waiting for patient data - need to think about mapping - in paper?
colin: that was me
michel: colin - do you need help with mapping?
colin: let me know how you get on
michel: just in paper at moment so let's migrate that to google doc
colin: OK
susie: PAPER
susie: goal to move paper into Word and make edits there on Monday
susie: Elgar - you'd volunteered to make SW box
elgar: have descriptions and definitions for about terms; too long
susie: anja was working on data sources; for ui we'll talk about 1. pubby and 2. what we'd like to see.
susie: chris says the veterans' association system is one that physicians like - bosse is leading on openvista (OS version of this)
susie: goal to make web client that looks like openvista. Eric Miller (attached) interested in doing a mockup
susie: bosse - what update?
bosse: have added section to paper; missing anja's contribution
susie: thinking of having a breakout group
susie: eric - anything to add
eric: getting bearings and finding out where assets live
bosse: eric - look at what we have in the document
susie: will invite eric to have access
bosse: one question - one of the challenges is identifiers
bosse: shouldn't we mention identifiers as a challenge?
susie: URIs will be in box and also text about SW approach, but no discussion about common URIs
susie: challenges of reconciling identifiers
susie: good progress; should be really good paper