Hello everyone,

In attendance we had:

UCSF - Rob Wynden, Gail Harden, Paul Norris and Hillari Allen

UPenn - Maggie Massary and Liat Shimoni

URochester - Marco Casale and William Digrazio

Here is a recap of today's call.  The first half of the meeting was about the Researcher Lifecycle paper.  The second half of the call was about a paper on our proposed extensions to the i2b2 schema in order to accommodate the Ontology Mapper.  We will be writing a paper for each of these two tasks.

Liat, Paul and Hillari will be focusing primarily on the Lifecycle paper but also attended the second half of the meeting as well.

Lifecycle Tasks:

Purpose: Write a paper that documents the likely interaction of a research with an Integrated Data Repository from project inception, the design and running of a study, publication and archiving the completed work.  What are the types of usages and queries our customers will generate?  What kinds of functionality would likely be useful to them?  How will they interact with the IDR and what sort of document artifacts will they be likely to generate?

1) Maggie has most graciously agreed to write the main document structure for the Researcher Lifecycle paper and she will serve as our First Author and moderator.  All of our work product will be reflected within this paper.

2) All parties agree to send Maggie any prior art that they had identified to help serve as a foundation for the paper.

3) Liat will also attempt to search the current literature on this topic.

Database Tasks:

Purpose: Write a paper that documents what database schema is required to use the Ontology Mapper.  Also specifically what extensions to i2b2 in particular are we building in order for i2b2 to use the Ontology Mapper?

1) Marco has most graciously agreed to assume the First Author role for the creation of our Ontology Mapper database design extension paper.  All of our work product will be reflected within this paper.

2) Rob will contact Stuart Turner at UCD to see if he can assist us with the absorption of his HL7 mapping data model into the database.

3) Rob will post a question on the i2b2 forum about this topic to see if others had also thought about schema extensions for custom UI work.

4) Questions raised:

    A) Are the 'like clauses' okay?  What impact would a table scan have on an ontology which is significantly larger than ICD-9?

    B) Do we extend the current concept table design or create a relational database design?  How about the hybrid option that William has suggested?

    C) Alternate encodings seem to be best represented as new rows within the main fact table.  I2b2 may already have a mechanism of associating a data encoding with other alternate encodings for the same data.  (We may want to contact Griffin Weber about that)

    D) What sort of data are we storing?  Is it just the metadata for the Data Discovery and Data Request UI and the HL7 mapping subscriptions?

ALSO, of some import but not covered on the call: UCSF has hired Prakash Lakshminarayanan to begin the physical design for this system.  This week Prakash is installing the Protégé tools and beginning to design Protégé's interaction with the Ontology of Mapping Relations.  I have also asked him to read up on creating Protégé Visualizations so that we can have a custom Protégé UI for Ontology Mapper to make creating new mappings as easy as possible.

Thanks again for a very productive call!

Rob