26 Jul 2012

See also: IRC log


matthias, JoanneLuciano


<matthias_samwald> scribenick: matthias

<JoanneLuciano> Matthias: done with prototype, wants to move to clinical decision support application

<JoanneLuciano> BobF: working on domain model for guidelines

<matthias_samwald> scribenick: JoanneLuciano

moving towards thinking about implementation

<matthias_samwald> (thanks Joanne)

and linking to a CDS repository

(clinical decision support)

currently focused on guidlines that are not computable -- to make to a structured format and link to a standardized representation for a DS report rule that would help the clinical guideline

has been looking at CDS but got sidetracked to look at provenance, which will be important for the instutions to track (for status and for tracking)

looking into w3c provenance working group

Matthias reports some overlap with HCLSIG, Joanne reports RPI expertise in the area and activity in the group.

Bob - how to addopt the prov model to support clinical descion support rules

evidence or provenance only.

BF wasn't going to try to model evidence at this point -- currenlty looking at two types of guidlines on PharmGKB:

<BobF> CPIC - clinical pharmacogenetics implementation consortium

<BobF> dutch pgx working group

aiming to standardize into a common model - representing content of the guidelines

Joanne: possibly overlap and opportunity for dovetailing with Drug Product Label and Drug Drug Interaction task group

Matthias: Spoke with Anita de Warrd re: work with Philips - looking to come up with a prototype for EHR where clinicans are shown relevant snipits

MS - voices concern re: scope if we get too far into provenance and evidence

how will info be entered and maintained

spreadsheets are a good way to collaborate with those users who are not computer experts

Bob - finish up model and take one or two guidelines, put in XML-schema that matches model and..

Bob exploring representations, for example, GEM - formalizing clinical guidelines / maybe able to be reused / how to represent and serialize the guildelines and get in position to seek funding. possibly thru a repository

Matthias: skeptical about using GEM and others because they are not easy to use and overly complex

{an asside ... this just out: We are very pleased to announce Last Call on 3 documents that define the core of the PROV family of specifications. These 3 documents focus on how to model provenance information for the purposes of interchange. - PROV-DM is a data model for provenance interchange - PROV-O is an OWL2 ontology mapping of PROV-DM to RDF - PROV-N

bottom line - something based on a spreadsheet would be useful; BF - will think about, interesting perspective

<matthias_samwald> http://www.snpedia.com/index.php/Criteria_syntax

<matthias_samwald> SNPedia criteria syntax

<michel> HGNC nomenclature : http://www.hgvs.org/mutnomen/recs.html#general


CONTEXT: currently this discussion is focusing on the use cases described on the HCLSIG CDS web page, specifically: formalize (use case 1) and integrate (use case 2)

<michel> star allele nomenclature described here - http://www.nature.com/clpt/journal/v82/n3/full/6100284a.html

most guidelines use a combination of ways to specify criteria (e.g. *1 or *2) other guidelines will specify "if one variant allele) - which then indicates which guideline to reference

<matthias_samwald> https://docs.google.com/spreadsheet/ccc?key=0AiGT-vnkGcoLdFFVMEdqcFdYaDFqS0xHTnlUT0N3cEE&hl=en_US&pli=1#gid=3

the data will evolve... and the interpretation of phenotypes are subjective (intermediate metabolized, poor metab) based on distrib of activity that overlap

phenotype based classification of rules is problematic, so unless doing an enzymatic test, you're going to be starting with genotype info, so BF thinking of ways to support he phenotype based, which are being used all the time, so wants to support with genotype based.

want more precise guidelines

objectives - easily maintainable by non-computer experts

Michel - adventures with OMI (?)

<matthias_samwald> OMIM

Michel wrote a parser for OMIM - use OMIM API to retireve OMIM records (better structured than download) - turned to RDF and can get access to SNP entries (all diseases that have links to DBSnp)

<michel> one example : http://bio2rdf.semanticscience.org:8009/describe/?url=http%3A%2F%2Fbio2rdf.org%2Fomim_resource%3A608547_allele_1

MS: how many of those entries are relevant for Pharmacogenomics?

Michel is composing a SPARQL query in real time to answer :-) (Michel is awesome)

And the answer is......

genes, phenotypes, of genes or phenotypes, predicate is variant in OMIM vocab. all variants that have a DBSnp entry and count them.

1767 have entries

<michel> query: SELECT count(distinct ?a) WHERE { ?a <http://bio2rdf.org/omim_vocabulary:variant> ?b . ?b <http://bio2rdf.org/omim_vocabulary:dbsnp> ?y . }

out of 23,000 (approx)

about 9 % coverage (roughly)

But the clinical reference can only be determined by PharmGKB and to see if there is an impact in pharmocogenomic or pharmocodynamic (free text in PharmGKB), can possibly annotate ourselves(?)

Drugbank, PharmGKB, OMIM need to be integrated to answer the questions

recently upgraded machine to: 192 GB RAM, so should be enough for now

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.136 (CVS log)
$Date: 2012/07/26 16:00:09 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.136  of Date: 2011/05/12 12:01:43  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Found ScribeNick: matthias
WARNING: No scribe lines found matching ScribeNick pattern: <matthias> ...
Found ScribeNick: JoanneLuciano
Inferring Scribes: matthias, JoanneLuciano
Scribes: matthias, JoanneLuciano
ScribeNicks: matthias, JoanneLuciano

WARNING: No "Topic:" lines found.

WARNING: No "Present: ... " found!
Possibly Present: BobF CONTEXT Joanne JoanneLuciano MS MacTed Matthias achille_zappa egonw ericP https matthias_samwald michel query scribenick
You can indicate people for the Present list like this:
        <dbooth> Present: dbooth jonathan mary
        <dbooth> Present+ amy

WARNING: No meeting title found!
You should specify the meeting title like this:
<dbooth> Meeting: Weekly Baking Club Meeting

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Got date from IRC log name: 26 Jul 2012
Guessing minutes URL: http://www.w3.org/2012/07/26-hcls-minutes.html
People with action items: 

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]