HCLS$$F2F$$BioRDFBIONT$BioRDF

Download Report

Transcript HCLS$$F2F$$BioRDFBIONT$BioRDF

BIONT-BIORDF Collaboration
Use Case
• Show me the location of receptors that bind to a ligand which is
a therapeutic agent in {Parkinson’s, Huntington’s} disease in
each of the dopaminergic neurons in the {pars compacta, pars
reticularis, substantia nigra}.
• Spans the following domains
–
–
–
–
Compartments of Neurons (Anatomy, Systems Physiology)
Receptors on Neurons (Protein, Molecular Biology)
Compounds that bind to Receptors (Pharmacology)
Ligands associated with disease (Clinical/Medical)
• Spans the following Data Sources
– Neuron DB (SenseLab), RDF
– KI DB, OWL
– PubChem, TBD
BIORDF Hackathon
• Data Sources:
– Cocodat (SenseLab)
– KiDB
• Alan shows demo (at the end)
Tasks for Demo
• Specify SPARQL queries against Parkinson’s Disease Ontology
• Annotate Data Sources with
• Ontological relationships, e.g., receptor located_in neuron
• SPARQL queries to populate the relationships
• Create “Merged Data Warehouse” containing a pre-integrated database.
• “Manually Decompose” SPARQL query against ontology into SPARQL sub
queries against individual data sources
• Merge RDF graphs
• User Interface?
• Update/refine Parkinson’s Disease Ontology
Architectural Alternatives
• “Data Integrated vs Accumulation Warehousing”
• “Mediator” approach
• Discuss Pros and Cons
• Demo: April 2007
• Report could describe:
– Approach for handling use case (as we implement it in the demo)
– Architectural Alternatives
– Value proposition of SW technologies for data integration, user interface and
distribution?
– Alternatives for semantic web based representation of data…
– Editorial Lead: Don Doherty
– Deadline: November 2006
Next Steps
• Prepare Note in the November time frame
• Propose joint teleconferences/wiki pages and sessions to map
out the tasks and set up a demo
• Determine the sophistication of the demo. To what extent
should we “implement” the mediation architecture ? Or is just a
simple hard coding of SPARQL queries enough ?
• Figure out potential clinical extensions based on Parkinson’s
guidelines identified by ACPP group
BIONT Report
• Ontology Modeling and Mapping Issues
– Possible Collaboration with ACPP
– First Draft: November 06
– Mapping across different ontologies such as:
•
•
•
•
•
•
Gene Ontology
MeSH
Neuro Names
Enzyme Commission
RIM
Galen
– Editor: Vipul Kashyap, Helen Chen?