HCLSIG/PharmaOntology/Meetings/2010-03-18 Conference Call

From W3C Wiki

Conference Details

  • Date of Call: Thursday March 18 2010
  • Time of Call: 12:00am - 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: Elgar
  • Scribes: Bosse, Elgar


Agenda

  • TMO
    • Data - Michel
    • Ontology - Elgar, Colin
    • Interface - Bosse
    • Outreach - Trish, Michel, Colin, Chris
  • Inclusion/exclusion criterion - Michel, Bosse
  • Other Outreach
    • UPenn Translational Medicine (ITMIT) - Chris
    • Other opportunites (AAAI)
  • Upcoming F2F
  • AOB


Minutes

Attendees: Bosse Andersson (B), Colin Batchelor (Co), Chris Domarew (Ch), Michel Dumontier (M), Elgar Pichler (E), Trish Whetzel (T)

Apologies: Susie Stephens (S)


TMO

- Ontology:
  • E:
    • noticed new version of TMO has been uploaded to TMO Google project
    • review of recent ontology work?
  • M:
    • went back to comments of JBI paper reviewers; fixed hanging terms, moved orphan classes as information content entities that define various contextual information (information about drug, patient, etc.)
  • Co:
    • information entities grouping is interesting, will take a closer look at TMO changes
  • E:
    • did the comments from mail exchange in December make it into the ontology?
    • thought also it was time for a second round of TMO editing;
    • wanted to address 3 things: 1. missing concepts?; 2. Cannot display IAO properties labels in an informative way in any ontology editor - IAO annotations rather difficult to figure out for anyone who looks at TMO; 3. How can we make mapping of concepts to other ontologies more dynamic? (i.e., actually enable a TMO to dynamically use part of another ontology if we point to class in it via xrefs; at the moment we just provide xrefs and for some classes a static link to another ontology)
  • T:
    • we should use SKOS properties for editorial notes
  • E:
    • there seem to be several choices for annotations: 1. IAO, 2. SKOS, 3. annotations in collaborative Protege, ...; IAO currently has usability issues; SKOS is definitely a choice - could Trish suggest when to use it?
  • Co:
    • agreed, problems with IAO
  • T:
    • could use SKOS, e.g., for prefLabel, altLabel (for synonyms); will compile a short list of properties to use from SKOS
    • ⇒ T: compile list of SKOS properties which could be used for TMO
- Data
  • M:
    • had discussion in February about mapping RxNorm - any progress?
  • E:
    • yes, have completed UMLS install (includes: all unrestricted sources (level 0) + SNOMED)
    • now can extract UMLS mappings between all those source vocabularies
    • could now extract RxNorm mappings to Alcohol and Other Drug Thesaurus, Metathesaurus FDA National Drug Code Directory, and Veterans Health Administration National Drug File
    • ⇒ E: provide RxNorm mappings to M for integration into TMO data triple store
- Inclusion/Exclusion Criteria
  • M:
    • had conference call last week with Bosse and Vassil (OntoText) and discussed how to extract inclusion/exclusion criteria from LinkedCT
  • B:
    • work also of interest for LOD project
  • M:
    • bjective: to be able to query over these "entities" and relationships
    • have developed script and RDFiser to download and transform ClinicalTrials.gov to RDF
- Interface
  • B:
    • presented in last TMO TC a schematic view of interface scenarios (TMO use scenarios) that were developed with Eric; Eric needs comments on these in order to better gauge interface requirements; any comments?
  • E:
    • if there are no comments now, we all should look at this and comment
    • ⇒ all: check out TMO use scenarios and send interface comments to Bosse


Outreach

- Report from AMIA2010 Summit on Translational Bioinformatics, CSHALS 2010
  • Ch, T:
  • Ch:
    • at conference last week met people from Mayo Clinic [should that be Marshfield Clinic?] and initiated a discussion around collaboration on user interaction.
    • they have established eHR system; data dumped into Access; ran into interface issues
    • will set up a call next week; will send out meeting information and try to coordinate so that interested TMO members can participate
    • ⇒ Ch: to set up call between Mayo/Marshfield Clinic folks and TMO group
    • met Vivien Marx who wrote article on TMO
  • E:
    • yes, Vivien wrote nice article on TMO in GenomeWeb after CSHALS2010
    • regarding eHRs: talked with Jonas Almeida and Lynn Vogel (both from MD Anderson Cancer Center) at CSHALS
    • they have done very nice work on eHR and clinical data analysis customized for MD Anderson; very impressive informatics organization there; see also Vogel's CHSHALS keynote talk
  • Ch:
    • work done in this area at University of Minnesota also very impressive
    • a lot of parallel activities out there, we can learn from them and demonstrate the strength in our work
    • smart way of making data anonymous is to shift the time so that the individual not can be identified
  • E:
    • seems we could follow different approaches in working with clinical data:
      • 1. get clinical data (very hard for US data because of privacy issues, might be easier to get data from countries other than US) and integrate with other TMO data, or
      • 2. leave data with owners behind firewalls (e.g. at hospitals) and strike deal with data owners to jointly analyze data with TMO tools
  • Ch:
    • critiscism of ontologies: don't strife forever to model everything; use existing systems; find something that works, apply it, don't reinvent wheel
    • ontologies well received and accepted, but goal should be to make them work so that patient care improves
  • T:
    • other interesting people presentations at conference:
      • Allan Halevi [?], Google: why not mine what you got?
      • Peter Kos, Harvard: interested in further discussions with TMO
      • Sherri De Coronado, NCI: worked on similar project with grid technology
- former JBI paper
  • Co:
    • will talk about this at next TC


Meeting/Communication Logistics

- HCLS F2F
- return to weekly TMO TC meetings
  • E:
    • so as not to lose momentum, how about return to weekly meetings as previously discussed at beginning of year?; had a productive push last year to get JBI paper done; would be good to go for something similar
    • how do people feel about this?
  • Co, Ch, M, E, T:
    • support weekly schedule
  • B:
    • would like to have goal to work towards, otherwise meetings can also just take time
  • E:
    • agreed; but seems we have enough action items for next weeks to necessitate going to weekly meetings; we seem to have general preference for weekly TC schedule
    • ⇒ S: to reinstall weekly TMO TC meetings, starting 2010/03/25
- TMO mailing list
  • T:
    • would be useful to have TMO only email group so that TMO notifications and updates don't have to be sent out to whole HCLS mailing list
  • all:
    • sounds good
  • T:
    • ⇒ T: to set up google mailing list for TMO


Action Items

  • ⇒ Elgar: provide links for meeting minutes
  • ⇒ Trish: compile list of SKOS properties which could be used for TMO
  • ⇒ Elgar: provide RxNorm mappings to Michel for integration into TMO data triple store
  • ⇒ all: check out TMO use scenarios and send interface comments to Bosse
  • ⇒ Chris: set up call between Mayo/Marshfield Clinic folks and TMO group
  • ⇒ all: sign up at F2F wiki page, if you plan to attend
  • ⇒ Susie: reinstall weekly TMO TC meetings, starting 2010/03/25
  • ⇒ Trish: set up google mailing list for TMO