HCLSIG/PharmaOntology/Meetings/2009-01-22 Conference Call

From W3C Wiki
Jump to: navigation, search

Conference Details

  • Date of Call: Thursday January 22, 2009
  • Time of Call: 12:00pm Eastern Daylight Time (EDT), 17:00 Greenwich Mean Time (GMT), 18:00 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: 42572 ("HCLS2").
  • IRC Channel: irc.w3.org port 6665 channel #HCLS2 (see W3C IRC page for details, or see Web IRC)
  • Duration: ~1h
  • Convener: Susie
  • Scribe: Christi

Agenda

  • Overview of proposed Pharma Ontology
  • Discussion of deliverables, scope, and next steps

Minutes

Attendees:

  • Sam – Boehringer Ingelheim w/ Ming Shu: text mining and data mining for Biology
  • Jackie – Amgen: works with Charlene Clayton, ontology projects
  • Christi Denney – Lilly: discovery data management, information integration team
  • Michel Dumontier – Carleton University: ontologies for pharmacogenomics data (http://dumontierlab.com/pdf/2008_HCLS_pharmacogenomics.pdf)
  • Chad – Microsoft: FAST search acquisition. Working with Amgen and others on search
  • Rob Frost – VectorC
  • Olivier Bodenreider – National Library of Medicine: research on UMLS
  • Claudia Schepers – Merck KGaA: KM for research and development
  • Jeremy Packer – Abbott Bioinformatics
  • Jennifer – Amgen
  • John Hart – Boehringer Ingelheim: Term lists for Target ID/Target Validation
  • John Hill – Boehringer Ingelheim: Term lists for Target ID/Target Validation
  • Nathan Siemers – Bristol-Myers Squibb: KM initiatives, biomarker, enterprise search
  • William Hayes – BiogenIdec: linking info w/ corporate ontology
  • Bosse Andersson – AstraZeneca: semantics
  • Joanne Luciano – MITRE: ontology and translational medicine and clinic integration; working with Harvard medical
  • Davide Zaccagnini – Language and Computing: natural language processing and search
  • Dan – Recombinant Data: builds data warehouses for pharma/medical clinical data
  • Elizabeth Wu – Alzheimer’s Research Forum: SWAN disease hypothesis
  • David – Pfizer: high-throughput generation, working on some ontologies
  • Susie Stephens – Open Innovation at Lilly: W3C HCLS co-chair
  • Ted Slater – Pfizer: semantic work, hypothesis generation with experimental data
  • Eric Neumann – Clinical Semantics
  • Holger Stenzhorn – Institute for Medical Biometry and Medical Informatics
  • M. Scott Marshall – University of Amsterdam, W3C HCLS co-chair
  • Elgar Pichler – AstraZeneca
  • Egon Willighagen

<Christi> Susie: exploratory call to see if this is an area for HCLS

<Christi> Susie: interested in finding out if there is sufficient value

<Christi> Susie: Lilly is moving from compound-centric to patient-centric development

<Christi> Susie: Important to bridge much more data for patients and link in more of the health care-related data

<Christi> Susie: Need concrete use cases to work toward

<Christi> Susie: Web site has questions that scientists are asking. Need to expand this list and then see what existing ontologies may meet our needs.

<Christi> Susie: There are gaps in the middle. Need to recognize where those are.

<Christi> Susie: Build the ontology and test it. Incorporate into an application.

<Christi> Susie: Goal of building something and testing it to help avoid over abstraction of what we want to achieve.

<Christi> Susie: Need timelines and goals to work toward, not just an academic exercise.

<Christi> Susie: Started some ontology work at Lilly. Have shared an ontology with HCLS previously, so may be able to share this.

<Christi> Susie: Ontology shouldn't mirror Lilly's process.

<Christi> Susie: Having a common model for pharma can help us all with collaborations.

<Christi> There are use cases on the HCLS page that we may want to review. The Parkinson’s use case breaks up the questions by role.

<Christi> Tailor the questions to particular roles who would be asking questions of the ontology.

<Christi> The HCLS use case link is above (http://esw.w3.org/topic/HCLSIGUseCases).

<Christi> David: Does this approach apply mainly to research or can we apply to other areas? (e.g. post market evaluation, FDA data)

<Christi> Susie: It is more than just Discovery that is interested in this approach.

<Christi> HCLS Use case site - http://esw.w3.org/topic/HCLSIGUseCases

<Christi> Olivier: approach caution due to size, but am in support of pharma collaboration in this area

<Christi> Nathan: Need to share data across trials, competitive intelligence, research. What domain is the core element use case?

<Christi> Susie: There is a lot of flexibility. Doesn't have to be in the discovery space.

<Christi> Susie: We can collect use cases and then prioritize them.

<Christi> Susie: Contribute 4-5 line paragraphs for areas of interest, then have a call where we narrow down. Focus on one use case at a time.

<Eric N> Would recommend asking more scientists to get involved in this effort in order to own it rather than doing it within an IT group.

<Christi> Susie: high-level bridging ontology that could connect various areas, rather than focused on a very narrow area.

<Christi> Joanne: Criteria for achieving goal?

<Christi> Susie: Need well-defined use case and scope

<Christi> Joanne: Need to identify the different intentions

<Christi> Susie: enabling people to ask questions across multiple silos of data

<Christi> David: need to solve the problem of un-curated resources or incomplete resources

<Christi> Susie: linked data community taking data on the web and mapping it to rdf.

<Christi> Susie: converting publicly available datasets to RDF. e.g. clinicaltrials.gov

<Christi> Susie: also looking at UniProt and others from NCBI

<michel> check out the bio2rdf slides @ ISMB: http://www.slideshare.net/fbelleau/bio2rdfismb2008

<Christi> David: integration is key

<Christi> Ted: recommends top-level ontology based on something already existing. lightweight ontologies idea would work for his group and hypothesis generation.

<Christi> Ted: post ontologies on the wiki page

<Christi> Dan: some basic items (e.g. cell line, clinical trial) are not defined. would be value in defining these.

  • Elgar is new here, and wonders about ontologies for drugs/medicine (I come from a cheminformatics angle)
  • Scott M tells Elgar - welcome to the leading edge. :) We can certainly use cheminformatics expertise when talking about compounds, drugs, structural homology, chemical properties, PSA, etc.

<Christi> Susie: Will create more space on the wiki page for documenting resources and use cases. We can get together in a couple of weeks and review.

  • Egon tells Scott M - I'm eager to join the show... been slowly moving to RDF for molecules (e.g. rdf.openmolecules.net) and now working on extracting molecules from DBPedia
  • Scott M to Egon - I've noticed your activity. Thanks for any contributions.
  • Egon tells Scott M - being in a drug discovery group now... linking such to disease, protein, whatever is quite relevant to what we plan to do this year

<Christi> Joanne: conference in cambridge for semantics in hcls.

<Joanne> C-SHALS: www.iscb.org/cshals2009/)

<Joanne> feb 25-27, cambridge ma

  • Scott M tells Egon to also check out LODD task force (Wed afternoon in NL).

<Christi> The next meeting will be on February 5 at 12 ET

Action Items:

  • Christi post the minutes to the web site
  • Susie to include details on C-SHALS conference in mail to the group
  • Susie add a wiki page for use cases
  • All contribute use cases
  • Susie add a wiki page for a list of existing ontologies and resources
  • All contribute existing ontologies
  • Susie announce the next call