HCLSIG/PharmaOntology/Meetings/2009-09-10 Conference Call

From W3C Wiki

Conference Details

  • Date of Call: Thursday September 10 2009
  • Time of Call: 11:00am - 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: 2h
  • Convener: Susie

Agenda

  • Refinement of Classes and Class Definitions
  • Use case polishing - Chris
  • F2F meeting - Susie
  • Progress on time line - Susie

Minutes

Attendees: Colin, Anja, Julia, Bosse, Christi, Trish, Elgar, Michel, Chris, Eric, Susie

Apologies: Elgar

<Susie> Susie: F2F

<Susie> Trish - likely to attend

<Susie> Christi - very unlikely to attend in person

<Susie> Bosse - Maybe able to attend

<Susie> Julia - Maybe able to attend

<Susie> Julia - Elgar may be able to attend

<Susie> Colin - Very unlikely I can attend in person

<Susie> Susie: Focus on next steps after the paper

<Susie> Susie: Everything is open, even wrapping up the task

<Susie> Colin: Ontologies need caring for, so unlikely we could wrap up the task

<Susie> Susie: Made comment to be somewhat provocative

<Susie> Joanne: It's likely the deadline for papers will be extended, so we may want to work on the paper during the F2F

<Susie> Susie: very good point. So let's have a flexible agenda, so we can easily work on the paper if the deadline is extended

<Susie> Susie: Use Cases

<Susie> Susie: Chris has worked to enhance the Chemogenomics and Pharmacogenomics Use Cases

<Susie> Susie: Susie will post onto wiki while Chris works out his login information

<Susie> Susie: Let's select 6 or so use cases during next weeks call to map to the ontology

<Susie> Trish: Want more than one ontology to make sure that we don't overfit it to a use case

<Susie> Susie: Let's also identify the one use case we want to pursue

<Susie> Julia: That would make Elgar happy

<Susie> Julia: Elgar also wants us to identify the data

<Susie> Colin: Seems like we're close to the point where we are going in circles with class definitions, so more of a focus on use cases would be good

<Susie> Susie: Will create a time line that we should work towards in order to make sure that we don't miss the deadline

<Susie> Susie: Classes

<Susie> Colin: Filled in class information

<Susie> Colin: Do we need 'ClinicalGuideline'

<Susie> Susie: Would be happy to just have 'ClinialProtocol'

<Susie> Joanne: Depends on use cases

<Susie> Susie: Use cases don't drill into these details a lot

<Susie> Christi: We should delete

<Susie> Colin: What about 'Clinician'

<Susie> Susie: Don't think we need it

<Susie> Christi: Maybe need specialities

<Susie> Christi: need role

<Susie> Colin: Like the idea of 'Expert'

<Susie> Chris: Isn't there something that lists all experts

<Susie> Susie: We're only looking for a minimal set of 30, so loose Clinicians

<Susie> Colin: Lets keep 'Expert' for now

<Susie> Colin: Company

<Susie> Susie: No

<Susie> Joanne: No

<Susie> Trish: Like company, but OK to go with concensus

<Susie> Colin: ConnectionTable

<Susie> Susie: Think it's too detailed

<Susie> Miche: Lots of in silico work

<Susie> Colin: Instinct is to keep it

<Susie> Colin: CostBenefitAnalysis

<Susie> Chris: anyone involved in this

<Susie> Christi: Something pharma looks at

<Susie> Chris: Rather important from clinical view

<Susie> Colin: Is this something we have data for

<Susie> Susie: Pharma interested in comparative effectiveness, disease population, patents

<Susie> Chris: Important in determining treatment

<Susie> Susie: Lets defer to Chime and Vipul

<Susie> Colin: Dosage

<Susie> Susie: Too detailed

<Susie> Christi: Too detailed

<Susie> Chris: CYP influences warfarin dosage

<Susie> Susie: Happy to cover dosage under ClinicalProtocol

<Susie> Michel: Don't think it should be under ClinicalProtocol

<Susie> Christi: See it as part of the treatment regiment

<Susie> Christi: Rather than ClinicalProtocol per se

<Susie> Colin: There isn't another ontology to defer 'Dosage' to

<Susie> Colin: Keep but specify as non-core

<Susie> Colin: Drug

<Susie> Colin: Covered by active ingredient, and pharma product

<Susie> Christi: Actually need as pharma product could be an insulin pen

<Susie> Colin: Effector

<Susie> Christi: Isn't a major piece, but is needed

<Susie> Colin: Shall we call it 'yes, but non-core'

<Susie> Colin: Finding

<Susie> Colin: Let's defer to OBI

<Susie> Susie: Need to include high level terms, but it seems too generic

<Susie> Chris: We've already compartmentalized everything, e.g. side effect, observation

<Susie> Chris: What would blood test go into

<Susie> Colin: Isn't it a sign, as it's something that's discovered by a physician

<Susie> Colin: Can't we use 'BiologicalMeasure'

<Susie> Colin: Generic

<Susie> Colin: Were going to refer to Elgar's hierarchy

<Susie> Chris: Would like to keep, as several pharma drugs become a generic

<Susie> Colin: Don't think it's our job to keep track of names

<Susie> Susie: Delete

<Susie> Joanne: Delete

<Susie> Colin: HealthOutcome

<Susie> Colin: Do we need 'healthoutcome' and 'outcome'

<Susie> Colin: Delete healthoutcome

<Susie> Colin: Hypothesis

<Susie> Susie: Keep

<Susie> Christi:

<Susie> yes

<Susie> colin: Institution

<Susie> Colin: keep

<Susie> Colin: Lead

<Susie> Susie: Lead is very similar to compound

<Susie> Colin: Do we want to differentiate between compounds we're pursuing and not

<Susie> Colin: Isn't the ontology only going to be used on public data

<Susie> Susie: No, internal too

<Susie> Christi: Want to use it internally too, and likely extend

<Susie> Colin: Delete

<Susie> Colin: Medical History

<Susie> Colin: This is of interest

<Susie> Colin: Molecular function

<Susie> Colin: Already have 'pathway' and 'moa'

<Susie> Susie: 'MolecularFunction' is more fine grained that pathway or moa

<Susie> Susie: Want as much biological detail as possible

<Susie> Susie: Would prefer 'MolecularFunction' to 'Effector'

<Susie> Colin: Could hand of Effector to Chebi or GO

<Susie> Colin: 3DMolecularStructure

<Susie> Michel:How does this differ to ConnectionTable

<Susie> Colin: ConnectionTable is 2D only

<Susie> colin: Rename ConnectionTable to MolecularStructure

<Susie> Colin: Then say no to 3DMolecularStructure

<Susie> Colin: Molecule

<Susie> Colin: Don't need in addition to compound

<Susie> Colin: Patient

<Susie> Chris: may want to broaden to animal studies

<Susie> Susie: individual may not be sick

<Susie> Colin: re-name to subject

<Susie> Joane: not convinced

<Susie> Joanne: Maybe we should have more of a focus on defintions

<Susie> Colin: Absolutely need definitions, and more important than names

<Susie> Colin: Prognosis and PatientPrognosis

<Susie> Christi: Looked at this as part of costbenefitpiece

<Susie> Colin: Prognosis is an outcome that may or maynot happen

<Susie> Colin: that's a yes

<Susie> Colin: Receptor

<Susie> Colin: That's very specific

<Susie> Susie: Delete

<Susie> Colin: Regulator

<Susie> Colin: is this the biological process, or the organization

<Susie> Colin: can't be both

<Susie> Christi: Have RegulatoryAuthority below

<Susie> Colin: Too detailed as a biological process

<Susie> Colin: Repressor

<Susie> Chris: Have we included the compound that can extend the life of an active ingredient and/or making it work

<Susie> Chris: Drugs are used synergistically more often these days

<Susie> Colin: Added 'ActiveIngredientStabilityRegulator'

<Susie> Chris: Works for me

<Susie> Colin: Repressor

<Susie> Susie: Delete

<Susie> Susie: ResponseBySubjectToDrug

<michel> i had previously suggested 'excipient' for the supporting material of a formulation. this covers the cases we discussed

<Susie> Colin: Have side effect and outcome

<Susie> Chris: Scratch

<Susie> Colin: Risk

<Susie> Susie: Could be risk of developing drugs

<Susie> Susie: Have lists of factors that make targets good or bad

<Susie> Susie: Likely we wouldn't share the list externally

<Susie> Chris: Could be patient risk

<Susie> Susie: subject risk is covered by side effects, prognosis, outcome

<Susie> Susie: Risk of drug discovery and development isn't covered

<Susie> Susie: But not sure it's needed

<Susie> Susie: Delete

<Susie> Colin: Safety

<Susie> Colin: Does it show up in the data

<Susie> chris: Safety is becoming more individualized

<Susie> Susie: Largely covered by outcome, prognosis, side-effect

<Susie> Christi: Safety is largely tied ot outcomes

<Susie> Colin: Let's keep for now

<Susie> Colin: Re-name 'Stratification' to 'SubsetofPopulation'

<Susie> Susie: Could encompass geography, e.g. phenotype, genetic, life style

<Susie> Colin: Yes

<Susie> Colin: Study

<Susie> Christi: Why did we say maybe

<Susie> Susie: Did compare study and experiment

<Susie> Susie: Let's keep

<Susie> Colin: Syndrome

<Susie> Susie: It was my term

<Susie> Susie: A catch all for diseases that aren't fully defined yet

<Susie> Chris: Sometimes a collection of diseases is a syndrome

<Susie> Colin: TherapyCostBenefit

<Susie> Susie: similar to costbenefitanalysis

<Susie> Colin: delete

<Susie> Colin: We have 1 maybe, roughly 30 no, and roughly 60 yes

<Susie> Colin: Should make the ontology the size it needs to be

<Susie> Colin: Will create figure of the ontology

<Susie> Colin: Will create in Protege in OWL

<Susie> Susie: Should we include Time

<ericP> http://www.w3.org/TR/2006/WD-owl-time-20060927/

<Susie> Susie: Time relates to patient stage, patent duration, etc

<Susie> Anja: Recommend connect to higher level representation

<Susie> Chris: Interesting seminar coming up in PA

<Susie> chris: Will send details

Retrieved from "https://www.w3.org/wiki/index.php?title=HCLSIG/PharmaOntology/Meetings/2009-09-10_Conference_Call&oldid=18304"