HCLSIG/PharmaOntology/Meetings/2010-04-09 Conference Call

From W3C Wiki

Minutes from todays clinical collaboration/UI Side Call as part of the TMO

Conor Bosse Chris Eric

Apologies form luke and from elgar –

Synopsis -

Goals :

To assess the current systems used by Marshfield and the other consortia at eMERGE to house, link, and query their clinical data. Once we are able to make a full assessment, we can offer in the form of advice or collaborative efforts (whatever is most comfortable and agreeable) a systematic approach toward semantically enabling data within the network for collective, federated query using an underlying ontology.

Baseline Questions :

1)Does Marshfield or others in eMERGE have a SPARQL endpoint?

2)How do they federate their data for query? If not done at present, is each partner within the network responsible for their own query and development of their own UI?

3)What is the standard relational schema that is used?

4)Is access alone what is currently used as a basic medium to serve as a platform for query? If not, then what is?

Comments :

Once we can assess the current level of interoperability, we can offer a solution in several easy, clear, and understandable steps:

Can we get to a SPARQL endpoint with one cohort of data (ie Marshfield) – this should be relatively easy to do Next, Can we then assist in moving data into RDF? OWL will then follow

Once we can provide a base level / foundation , this will serve as a mechanism for implementation that luke can then easily show to his group, then roll out to the other cohorts -

This will serve as a prototype model that will demonstrate concrete benefit in a clinical environment the technologies we readily discuss – and to date only have virtual demos to convey importance of a larger roll out 1)Data conversion 2)Data federation 3)RDF – most important 4)Linkability 5)Semantic capability 6)Easy development of a UI – once the data is in the right form! 7)Application of the TMO or other ontologies (access)

“We need to start to crawl then walk then run”


The two Biggest steps : Getting to a SPARQL endpoint Getting data into RDF

We need form luke : (In broad terms – for baseline assessment) His groups goals Marshfield’s goals eMERGE goals


Confidentiality issues : For reassurance to all parties If we provide Luke with information and tools to apply to his group’s data, and the support to implement them, we will be able to proceed through a logical sequence, and will effectively produce a real live working model using patient data We do not want to gain access to this data and respect the fact that it is confidential and not to be distributed or used by any members of our team

Rather, we would like to volunteer to enhance current projects the group may be working on at present to help forward all parties endpoint goals with respect to translational research