HCLSIG BioRDF Subgroup/Meetings/2010/01-25 Conference Call
- Date of Call: Monday January 25, 2010
- Time of Call: 11:00 am Eastern Time
- 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: 4257 ("HCLS")
- IRC Channel: irc.w3.org port 6665 channel #HCLS (see W3C IRC page for details, or see Web IRC), Quick Start: Use http://www.mibbit.com/chat/?server=irc.w3.org:6665&channel=%23hcls for IRC access.
- Duration: ~1 hour
- Frequency: bi-weekly
- Convener: Kei Cheung
- Scribe: Kei Cheung
Matt Cherian, Kei Cheung, Helena Deus, Don Doherty, Eric Prud'hommeaux, Matthias Samwald, Scott Marshall
- Introduction [Kei]
- Microarray use case -- exploring the RDF structure [All]
- Applicatios/demos [All]
<ericP> use cases for rdb2rdf:
<ericP> gene lists
<ericP> mscottm: many gene lists are not machine readable
<ericP> tabular structure can come from excel ericP welcome mcherian
<mcherian> thank you
<mcherian> does the telecon usually start at 1000 EST or 1100 EST?
<LenaDeus> kei: eric suggests using RDB to SPARQL query structure
<kei> matt: 2nd year grad. student at mit
<kei> matt: next to w3c office
<ericP> topic: RDF structure for gene lists
<kei> eric, can you scribe for Lena
<ericP> LenaDeus: considering annotating the gene lists from the mircoarray use case
<ericP> kei: you have an example of links to a specific gene card
<ericP> ... can we make that link broader, so folks can link other sources?
<ericP> ericP: depends on operational semantics
<ericP> ericP: is this typically in a spreadsheet?
<mscottm> Scott: We need to provide a gene list to Shared Na
<mscottm> .. Shared Names so that SN can arrive at an agreement about how the URI's will look
<LenaDeus> scott: shared names - given a gene list, which names we should use?
<LenaDeus> scott: how to approach the "/" vs "#" issue?
<LenaDeus> eric: motivation from both sides;
<mscottm> ericP: Another issue is that the slashes proposed by Shared Names are syntactically incompatible with the hash (#) approach
<LenaDeus> eric: shared names uses an "architecturally pure" approach
<LenaDeus> scott: how much data using hash vs dash?
<LenaDeus> eric: what are the relative costs of the two? the trade-off needs to be determined so that we can take a stand
<LenaDeus> eric: a variety of gene identifier that may be used (and that's the goal of shared names)
<LenaDeus> ericP: how to handle RDB to RDF mapping?
<LenaDeus> kei: mapping using the schema
<LenaDeus> kei: graph mapping rather than identifier mapping;
<LenaDeus> eric: is there a use case to use the federation of this with other data structure?
<LenaDeus> scott: taken a set of gene and looking at its function, the expression value may not mean much
<LenaDeus> kei: in some cases we want to look at over expressed genes, other cases we may want to compare the expression values
<LenaDeus> scott: the favorite gene approach - if a certain gene is know n to be involved in a phenomena, then it can be included in multiple studies
<LenaDeus> kei: next steps - start with the gene lists; learn how to convert it from excel to rdf
<LenaDeus> kei: how to demo?
<ericP> LenaDeus: two ideas for linking gene lists to experiements:
<ericP> ... .. simply say <experiementX> :generated <gene list Y> .
<ericP> ... .. include process (workflow) which led to the gene list
<LenaDeus> don: iphone application; were certain genes from the gene list already known to be involved in problematic pathways
<mscottm> Checkout: http://www.biomedcentral.com/1471-2105/10/S2/S3
<LenaDeus> scott: Rubin paper presents mapping genes to neurological regions
<LenaDeus> scott: mapping it eventually to an mri scanning would be useful to understand the process
<LenaDeus> mscottm: how to link with them
<ericP> how about links to myexperiment?
<mscottm> ACTION: Scott will contact Rubin et al to explore possible overlap
<mscottm> have to go to TERM call - bye!
<LenaDeus> action items: kei to add the spreadsheet gene lists; LenaDeus to create the gene list rdf structure and share it through a wiki page; consult with Satya and Jun how to connect it to experiment description (provenir) and see what metadata needs to be captured at the experiment level.