W3C

HCLS

12 Jun 2008

See also: IRC log

Attendees

Present
Chime, Susie, Scott, EricP, +~25 folks
Regrets
Chair
Susie Stephends
Scribe
ericP

Contents


Introduction to Co-chairs and Team contact - Chime, Scott, Susie

Susie: principal research scientist at Eli Lilly
... involved in RDF since 2002
... chaired BioRDF

chimezie: lead architect on SemanticDB at Cleveland Clinic

mscottm: postdoc researcher at Univ AMS

mscottm: work with lab for E-Science on knowledge management/capture
... involted with semweb
... compsci background, who has worked with biologists for a while

<Susie> Eric Prud'hommeaux sem web since 98, and team contact at W3C

<Susie> Happy to help with questions

Susie: encourage folks to post their intros on the wiki

Group Structure - Scott

mscottm: in 2002 there was a big meeting where people aired what they wanted out of the SemWeb in the areas of health care and life sciences

mscottm: an ambitious group, focused on data integration in health care, life science, biological modeling, tranlsational medicine

mscottm: crosses many databases

mscottm: first incarnation of the group started several task forces around areas of work
... in this incarnation, we expect to have several projects with weekly calls
... continuing from last charter, we have BioRDF task force and Clinical Observations Interoperability
... would like to gather project descriptions
... face to face meeting: 20-21Oct during the W3C Technical Plenary

mscottm: meeting is in Mandelieu

ericP: How do you manage a phone call with 40 or 50 people? We use IRC and some derivative of Robert's Rules of Order.
... we have a web-based IRC client which is available to Members and Invited Experts (more later)

<eneumann> Anyone except Susie from pharma on this call?

<mscottm> Therese Vachon from Novartis might show up when another TC is done.

<Zakim> ericP, you wanted to say something clever

ericP: there are some other tricks for managing agenda and that sort of thing.

<AdrianP> Hi, just joined

<mscottm> Hi Adrian

ericP: The W3C site is backed by a revision control system called CVS.
... A lot of these documents are edited by elected editors, responding to comments in publc mailing lists.
... Wiki's have their own form of concensus (most prolific wins).
... So W3C also has wiki space for things which invite that group dynamic.
... W3C is a member-based industry consortium
... that is, we invite participation from folks appointed by representatives of the various members.
... for this group, we'll have lots of invited experts

susie: note the email alias team-hcls-chairs

<mscottm> team-hcls-chairs@w3.org

Participation Expectations - Susie

susie: work expectations:
... .. participation is public, but we expect the majority will be members of W3C
... .. non-members can explore the IE path

<mscottm> IE = "Invited Expert"

susie: .. better to be a member, access to more resources
... .. we'll be pruning the mail list shortly to include members and IEs
... .. would recommend assigning a minimum of 10% time to the group
... .. appreciate that work preempts
... .. there will be less exploration than the first HCLS group; we expect the participants to be able to work on deliverabless
... .. deliverables will include conferences and papers
... .. meeting records will document action items, decisions, ...
... commitments:
... .. it's important to be clear when you can do something
... .. also important to respect other participants at all levels
... .. [ref Postel's Law]

-> http://en.wikipedia.org/wiki/Postel's_law Postel's Law

susie: i expect that following these guidelines will make a productive group

Charter Overview - Chime

chimezie: Charter defines the mandiate and scope of an interest group
... scope, goals, deliverables, liasons
... goals:
... .. 1st: advocate, develop tools, and suporting documents for SemWeb in health care, life science
... .. document use cases
... use cases useful for defining what we need to do and when we've succeeded
... will develop implementations that meet these use cases
... develop high-level vocabularies
... diseminate this info to govornments, other bodies...
... charter lays out separation of the primary domains of this group: biology, health care, trans therap
... primary value proposition is in integrating diverse databases
... Trans Therapeutics: working on "bench to bedside": preclinical trial data to affect patient medication
... health care
... coordiation with other groups:
... .. many of the standards in this area are sill develping

<AdrianP> would propose a liaison with W3C RIF

chimezie: another group to liase with is RIF

RIF WG

<AdrianP> yes, we are working on such use cases

<AdrianP> e.g. http://www.w3.org/2005/rules/wiki/UCR#Ruleset_Integration_for_Medical_Decision_Support

chimezie: relevent recently closed groups: SPARQL, GRDDL
... this IG is charterd for 3 years (starting 1 June)
... want to work with HL7, CDISC, those managing SNOMED -GT

eneumann: lots of nearby activities that aren't SemWeb-based
... the folks working BRIDG wanted to get more involved
... BRIDG working with CDISC, HL7, National Cancer Instituted
... they are now aware that there is an OWL version of BRIDG

chimezie: how does these liasons work
... we often invite them to speak to tasks forces

marco: how open are you to new use cases?
... if you work 10% of your time, it needs to be in sync with your interests

Proposed Projects - Chime, Scott, Susie

susie: acknoweldging the need for synergy beteen HCLS participation and day jobs, we've selected some tasks

<chimezie> Project Ideas

susie: we ask your advice on these tasks and invite new proposals
... if there projects that you are working on, you may enter it in the wiki page (Project Ideas)

<AdrianP> I already started to add to this projects

susie: alternatively, you can send email to the public-semweb-lifesci@w3.org list, or to the team-hcls-chairs@w3.org if you just want input from the chairs

tx AdrianP

<mscottm> http://esw.w3.org/topic/HCLS/ClinicalObservationsInteroperability

chimezie: Clinical Osbervations Interoperability task force (ongoing) lead by Vipul
... re-purposing existing clinical models and developing new ones
... lots of recent activity

susie: Vipul will give us a 20 minute overview during the next HCLS call (2 weeks from today)

chimezie: we encourage SNOMED to distribute in OWL

<mscottm> http://esw.w3.org/topic/HCLSIG/Project_Ideas/DecisionSupportRuleLanguages

<mscottm> oops - this one: http://esw.w3.org/topic/HCLSIG/Project_Ideas/DecisionSupportRuleLanguages

chimezie: ISTC is now planning to distribute as a secondary deployment
... RIF is a language for exchanging rules
... we can give them use cases in the area of Clinical Decision Support
... we can see how well these rules can support these use cases
... [task] identify archetypal data that comes up a lot in health care scenarios to push for standards in those small areas of KR

AdrianP: [editor of the RIF UC&R document] I can easily add your use cases to the UC&R doc

susie: CDISC creates clinical interchange standards approved by the FDA

<mscottm> http://esw.w3.org/topic/HCLSIG/Project_Ideas/CDISC

susie: [task] work with CDISC to create ontology of data elements withing the SDTM
... [task] (similar to drug safety and efficacy) labeling and modeling biomarkers

<mscottm> http://esw.w3.org/topic/HCLSIG/Project_Ideas/ClinicalTrials

<AdrianP> I'm also working on several use cases which exploit rules in HCLS

susie: [task] clinicaltrials.gov documents clinical trials around the world
... lillytrials.com includes lilly trials
... FDA publishes the Orange Book
... sometimes the data is available in HTML
... sometimes it's all trapped in PDFs
... project would involve picking a therapeutic area to mining the data in those papers
... [task] diseasome written up in nature, new england journal of medicine
... barabassy interested in working with us to publish diseasome in RDF/OWL
... [task] how can SIOC benifit the scientific/medical communities?

<mscottm> http://esw.w3.org/topic/HCLSIG/Project_Ideas/Diseasome

susie: this has become a big focus area for lots of folks

<mscottm> http://esw.w3.org/topic/HCLSIG/Project_Ideas/SIOC

susie: harvard just won a large grant for this
... create a graph of who's publishing with whom

<mscottm> http://esw.w3.org/topic/HCLSIG/Project_Ideas/PCMR

susie: [task] (less certain how to persue this just now) personally managed health care records
... clinical trials subjects get their genome scanned, but don't keep that information
... both google and microsoft are working on storage for this area
... we can be ahead of the curve (or leading edge) on this

mscottm: [covering life science/bio projects]
... some of these could form basis of projects that have already be defined

<mscottm> http://esw.w3.org/topic/HCLSIG/Project_Ideas/KB

mscottm: my top interest is enchancing the HCLS Knowledge Base
... could include adding diseasome data from barabassy
... would this find genetic reasons for diseases that are related to each other'
... [project] add chemical structures (KB currently biased towards neurology)
... can add other things like anatomy and physiology
... [task] distribution of the KB across multiple locations
... fairly challenging, people ask "can it be distributed?"

<AdrianP> we could probably host another triple store here at the Biotec Center in Dresden

mscottm: require nitty gritty work on query engines

<AdrianP> or a verbalized language, a controlled english interface

mscottm: [task] enhance the query interface to make it easier to compose queries without having to know all the namespaces

mscottm: similarly to getting to know the contents of a KB requires the same work

<mscottm> http://www.w3.org/TR/2008/NOTE-hcls-senselab-20080604/

<kidehen_> ericP: Does a query interface exist?

<ericP> kidehen_, only have SPARQL endpoints so far

<kidehen_> ericP: is the a link to something I can see?

<kidehen_> ericP: s/the/there

<mscottm> http://esw.w3.org/topic/HCLSIG/Project_Ideas/eScience

<matthiassamwald> kidehen_, some of us have built UIs that access parts of the Knowledge Base, such as Entrez Neuron: http://neuroweb3.med.yale.edu:8090/

<ericP> kidehen_, i think it's in the hcls-kb doc

mscottm: [task] add e-science to our work, including grid computing

mscottm: web services are a way for us to collaborate
... you can use Taverna (SP?) to compose workflows out of multiple web services

<mscottm> http://myexperiment.org/

mscottm: sharing workflows is getting easier
... you can trigger web services by clicking on something on a web page
... should look for synergy between these projects and other HCLS projects

<AdrianP> there is a synergy with the Rule Responder project, which allows to describe rule-based workflows

mscottm: e.g. exporting diseasome in RDF would be usefully shared
... e-science allows you to, e.g. export text-mined medline in RDF -- surprisingly in-reach

<matthiassamwald> shameless self-promotion of a prototype RDF/OWL wrapper for a text mining service: http://whatizit.neurocommons.org/

mscottm: main questions are "how do you wnat to represent the info, and how would you clean it up"

<mscottm> http://esw.w3.org/topic/HCLSIG/Project_Ideas/SemanticWikis

mscottm: also want to explore the use of semantic wikis

<dingl> we may look at some work on semantic media wiki at RPI

<dingl> http://tw.rpi.edu/wiki/index.php/Main_Page

Q&A

AdrianP: how do we get folks to work on a project?

susie: document the project, and if there's sufficient interest, off you go
... the lead person will pick task force meeting times, and report during the HCLS calls
... these projects will have indepencence and their own lifespan
... expect most to be done within a year

mscottm: if you have project ideas, try to keep in mind that synergy with other projects helps your goal
... helps to aim for a particlar demo/implementation

kei: could expand the KB with, e.g. herbal medicine
... china has dbs about these herbs

ericP: I expect that Huajun will be spearheading something like that

<Huajun> yes, I am working with kei together on this project actually

susie: western pharmaceuticals are interested in easter meds

<matthiassamwald> http://richard.cyganiak.de/2007/10/lod/

matthiassamwald: would be great aim to turn our projects into something that creates a coherent datastructure
... many independent participants in linked open data follow some rules to create an aggregated project

susie: reminds me that we'd like projects with common threads through the different silos
... don't want isolated projects; want to bridge them all together
... don't expect they'll all link together right from the start, but if you use e.g. the same compound as an example, we can link them together

Next call: two weeks June 26

susie: tx to all participants, tx for your participation

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.133 (CVS log)
$Date: 2008/06/12 18:53:34 $