See also: IRC log
246733 ("BIORDF")
I'll try again
<samwaldmatthias> i am ipcaller
<Susie> Olivier sends his regrets
<ericP> +004631776aaaa is Kirsten
BB: VK - review Action Items
<ericP> Scribenick: BillBug
VK: review Use Case (AD & PD) and look at demo script in more detail
SS: also
- review options for ISMB
- review JB SemWebTech pros
- review demo practical issues
VK: go SS
<vipul> Agenda
<vipul> Options for presenting at ISMB - Susie
SS: review ISMB presentation options
- bypass writing full-paper (time too short)
- Barb Bryant (ISMB chair) presented options
* send BMC Bioinform paper to highlights track
* PLoS track
* SIGs?
* organize Birds of Feather discussion (many BoF at ISMB)
VK: prioritize?
<alanr> Bill, you scribing?
Can I pass the baton to you, Alan?
<alanr> yes
SS
- 1) send BMC Bioinform
2) description of demo
3) BoF
4) SIG participation
5) PLoS Track
<scribe> done scribing
<samwaldmatthias> 21 - 25 of july
<vipul> Core Semantic Web benefits to highlight in the demo - John
<alanr> Advantages: Knowledge Representation
<alanr> 2) Knowledge integration
<alanr> - make sure consistent
<alanr> - deduce additional knowledge
<alanr> - in the case of OWL/RDF class hierarchy and class membership that is not explicitly declared
<alanr> - combine two "information resources" can use ontology to harmonize two resources
<alanr> - using rdms ontology becomes queries, new tables, possible constraints, but not well defined.
<alanr> - as opposed to OWL/RDF
<alanr> Problem: Pages on wiki immutable.
<alanr> ACTION: EricP to investigate problems with wiki [recorded in http://www.w3.org/2007/02/19-biordf-minutes.html#action01]
<alanr> Kei - multilayer representation, starting with annotation layer
<alanr> many sites are not structured. Adding annotation to sites would help with resource discovery
<vipul> q
<alanr> But: Standard vocabulary needed for this sort of annotation
<alanr> Kei - next layer: Integration
<alanr> allow users to interrelate concepts in meaningful ways
<alanr> in rdb: can supply foreign key, primary key but semantics has to be external - hard to discover relationship inside the rdb
<mscottm> agree with Kei - need to distinguish between domain semantics and data structure semantics.
<alanr> Beauty of semantic web is that it is describeed so machine can get at it.
<alanr> Third layer: Data available in OWL/RDF
<alanr> not only raw data - may not be that informative. Bill suggested distilled knowledge
<alanr> easier to integrate using SW technology.
<vipul> ack
<alanr> Bill, John: What pages were immutable?
<alanr> BillBug: Key thing: Semantics. Of minimal use unless we tie into shared formal semantic structures that people are already using.
<alanr> From controlled vocabularies to OBO Foundry.
<alanr> BillBug: Third dimension of Kei's is most important. Distilled knowledge resources *not* with data. Rather people mix them up.
<alanr> e.g. if you do text mining but don't do it in terms defined by these already extant ontologies then you lose the value
<Zakim> ericP, you wanted to advise "semantics of RDBs not expressed in SQL and not globally distinguishable". saying there are no semantics my annoy RDB folks
<alanr> Semantics are not global - can't move from one RDB to another
<alanr> for a good time: sw dot neurocommons dot org slash map.html <don't circulate for now>
<alanr> John: Difference between prolog and OWL-DL. Decidable and computable. Compare to prolog/uml
here, here - JB!
<alanr> June: Sent mail to Don re: use case
<alanr> From EricN: re: clinical
<alanr> Alan,
<alanr> Here is a draft for a possible clinical angle to the HCLS demo...
<alanr> 1. Extend PD/AD demo to include clinical study data from patients, using DSE's draft RDF model based on CDISC's SDTM standard:
<alanr> http://esw.w3.org/topic/HCLSIG/Drug_Safety_and_Efficacy/SDTM/Example_N3#Clinical_Data
<alanr> 2. Compile patient data for either:
<alanr> a. disease progression - base on symptomatic terminology such as EOAD or MCI for AD, and tremor, hyperkinetic dysarthia, or bradykinesia for PD
<alanr> or b. new drug trial - dosing, symptomatic response patterns over time + any observed adverse events (AE)
<alanr> 3. If drug trial, then links to PubChem drug information would be useful: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?CMD=search&DB=pccompound
<alanr> - for AD: Cognex, http://pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=2723754 (images- http://pubchem.ncbi.nlm.nih.gov/image/structurefly.cgi?cid=2723754)
<vipul> http://esw.w3.org/topic/HCLS/HCLSIG_DemoHomePage_HCLSIG_Demo
<alanr> - for PD: Cogentin, http://pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=441352
<alanr> or Piribedil, http://pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=198284
<alanr> 4. Use either set of data to show trends using a simple graph visualization (generic table to graph tool?)
<vipul> Maybe link it through from the demo wiki page?
<alanr> June discusses AD/PD Use case: Large, not intended to do all - more to pick from for demo.
<alanr> Layout: AD on left, data sources, Parallel PD use case. Some blank because they haven't been able to figure out parallels yet
<alanr> BillBug: Working on way of expressing specific assertions in narrow area , to use PATO to use phenotypic qualities.
<alanr> discussion at http://sourceforge.net/mailarchive/forum.php?thread_id=31662008&forum_id=35770
<alanr> BillBug: Means to move from June's description to formal representation
<alanr> vipul: Bill plz identify specific use cases he is representing. Bill: Linked to PD use case as existed in Sept/Oct
<alanr> is useful!
<alanr> Keep going!
<alanr> (Bill asks if people think this is useful - so he can decide how much effort to put into it)
<Don> Everyone...I just received an email back that the Use Case document exceeded the permitted size on the list.
<alanr> ACTION: Check feasibility of Bill's approach. All check out his examples. [recorded in http://www.w3.org/2007/02/19-biordf-minutes.html#action02]
<vipul> ACTION: Check feasibility of Bill's approach for focusing the use case [recorded in http://www.w3.org/2007/02/19-biordf-minutes.html#action03]
http://esw.w3.org/topic/HCLS/OntologyTaskForce/SeedOntology/SeedOntologyDetailedFollowup
<alanr> Action 2 is Don_Doherty
<alanr> June - initial thoughts on where to show benefits re: use case. Would like to integrate 3 databases. 1) SWAN 2) Genomic, e.g. expression. 3) NeuronDB and associated databases
<ericP> ESW wiki has account creation temporarily turned off due to spamming prob. i wonder if we can temporarily temporarily turn it on so folks can create their accounts
<alanr> K channel story, interferon gamma story. Too much to fold into one scenario
<alanr> Ask whether polymorphisms in channels associated with disease risks
<Don> Okay...people should be receiving the use case
<alanr> Got document.
<alanr> June: Mouse models/ does SNP exist in human.
<alanr> alan - just read about K channel SNP associated with deafness.
<alanr> Bill - idea about immunotherapy link. Bill could you plz elaborate on this to mailing list?
<Don> June...could we talk more about the Use Case on tomorrow's BIONT call?
<vipul> ACTION: Connect the benefits to the use case [recorded in http://www.w3.org/2007/02/19-biordf-minutes.html#action04]
<alanr> hearing loss discussion - http://sourceforge.net/mailarchive/forum.php?thread_id=31646450&forum_id=44607
<alanr> sw dot neurocommons dot org slash map.html
<alanr> ACTION: Susie to poll group on meeting before next SW Gathering [recorded in http://www.w3.org/2007/02/19-biordf-minutes.html#action05]
<ericP> hi all. account create is turned back on on the wiki
<ericP> but i may be telling you too late
<alanr> eric will you handle minutes generation please?
<alanr> rrsagent set minutes world-readable
<alanr> rrsagent set log world-readable
This is scribe.perl Revision: 1.127 of Date: 2005/08/16 15:12:03 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Found ScribeNick: BillBug Inferring Scribes: BillBug WARNING: No "Topic:" lines found. WARNING: No "Present: ... " found! Possibly Present: Advantages Alan_Ruttenberg BB BillBug Bill_Bug But Don Don_Doherty IPcaller John JohnBarkley June June_Kinoshita Kei_Cheung Kirsten Layout MIT-G346 P0 Problem SS Scott_Marshall Scribenick Susie VK Vipul_Kashyap aabb aacc alanr ericP mscottm samwaldmatthias vipul You can indicate people for the Present list like this: <dbooth> Present: dbooth jonathan mary <dbooth> Present+ amy WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth Got date from IRC log name: 19 Feb 2007 Guessing minutes URL: http://www.w3.org/2007/02/19-biordf-minutes.html People with action items: check connect ericp susie WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option. WARNING: No "Topic: ..." lines found! Resulting HTML may have an empty (invalid) <ol>...</ol>. Explanation: "Topic: ..." lines are used to indicate the start of new discussion topics or agenda items, such as: <dbooth> Topic: Review of Amy's report[End of scribe.perl diagnostic output]