SDIWG:Meeting Minutes 20060317

Home < SDIWG:Meeting Minutes 20060317

*********** HAPPY SAINT PATRICK'S DAY EVERYONE ***********

Agenda: Software and Data Integration Working Group (SDIWG)

Friday March 17, 2006: 2:30 -- 3:30 PM Eastern Time

Next SDIWG Friday April 21, 2:30 PM –- 3:30 PM ET: Pls contact Peter Lyster for information lysterp@mail.nih.gov

Attendees: Peter Lyster, David States, Mark Musen, Suzanna Lewis, Dan Rubin, Ivo Dinov, Bill Lorensen, Michael Sherman, Zak Kohane, Susanne Churchill, David Berkowicz, Karen Skinner, Jennifer Couch, German Cavelier, Valintina Di Fransesco, Zohara Cohen, Jody Sachs, Jon Dugan.
Note taker: Aris Floratos

[Suggested order of note takers for future meetings: Floratos (this one); Rubin; Jags; Dinov; Chueh; Sherman; Lorensen]

Minutes

Peter Lyster (PL): introduce agenda and invite Symbios people to provide an overview of their home-grown Yellow Pages solution (Simibiome).

John Dugan, Michael Sherman (Symbios): Provided a quick demo of Simbiome, showing the information captured for each entry in the yellow pages as well as the keyword search interface. Mentioned they have also developed a document comparing their solution to IATR. Explained that Simbiome is targeting the internal Simbios community and it was not developed with the intention of serving as a repository for a larger user base (although it might).

(Various): Does Symbiome use controlled vocabularies for any of the tool descriptors/fields or even for the field names themselves? Could author names, etc come from a list of keywords?

Symbios: At present Symbiome is not using keywords, it is rather free-text based. The objective was to create a system that people can use imediately to enter useful information; controlled vocabularies could be introduced at a later time, when the system had sufficient content that could be used to deduce such vocabularies. Symbiome allows the assignment of a "private" designation to a field, indicating that the data for the field should not be made public. Symbiome at present has about 75 entries, many more expected when it is opened to the entire Stanford/Symbios community.

David Berkowicz(DB): How is data entered in the system? Web form? Formatted data? Is there an RSS feed version of the repository data that is automatically updated when a Symbiome entry is added/edited/deleted?

Symbios: For now the Symbiome team solicits from tools authors information about tools in e-mails which they then enter in the system themselves. In the future other modes (Web forms etc) are possible.

David States (DS): What are the expectations from the authors that register their tools with Symbiome?

Symbios: (1) enter information that is correct, (2) contact person, (3) web site for the tool.

Bill Lorensen (BL): It would be convenient if there was a hub where to enter data and then an automated way to push that data in either the Symbiome or the IATR format.

Aris Floratos (AF): It would be convenient if there was programmatic access to the data.

Symbios: Many of these issues can be addressed at a later implementation.

PL: Invited Zak Kohane to comment on discussion and introduce the I2B2 efforts on developing a Yellow pages system.

Zak Kohane (ZK): Feels that we effort should focus on addressing the simplest issues first. Let's get the various software tools entered into a Yellow Pages repository of some sort and we can then start worrying about some of the more advanced fatures suggested. As a co-chair of the SDIWG "Scientific Ontologies" working group he intends to reach out to the various Centers to investigate if there are domain areas where the Centers can agree on a set of ontologies to use.

DB: Described how I2B2 is experimenting with the of use Maven (a tool coming out of the Apache project) to create comprehensive, web accessible information about various projects. Provided pointer to www.codehouse.com, a site hosting several projects which use Maven as their project management tool. Described how Maven provides menchanisms to create a project site by parsing documentations embedded in source code as well XML-formatted data conforming to the Maven project data model. Use of these mechanisms allows Maven-generated sites to remain up to date with the underlying development efforts.

PL: Maven sounds like a tool that goes beyond the "low hanging" fruit of creating yellow page services.

ZK: Suggest that the most important thing at this point is for each center to register their tools in a repository, regardless of which yellow pages technology and data model is used. Rather than wasting time until we all arrive to an common standard let everyone develop their yellow pages using whatever they want and we can worry later on harmonizing.

(Various): some discussion about pros and cons of centers proceeding using each their own yellow pages technology versus a consensus model.

David Kennedy (DK): Offered to post the IATR paper and schema on the SDIWG site.

AF/BL: MAGNet and NA-MIC are using IATR to register their tools.

PL: Presented one more yellow pages schema developed by folks at NIH. Valentina Di Francesco (affiliated with that effort) and Bill Lorensen commented on that schema's similarity to the IATR.

Karen Skinner (KS): Turned the attention of the group to the agenda item that was calling for the creation of a Charter/Objectives statement for the Yellow Pages and Scientific Ontologies working groups.

PL: Invited the Dan Rubin and Ivo Dinov to comment on the objectives of the Yellow Pages WG.

Ivo Dinov (ID): There are at least 2 types of artifacts that are candidates for registration in a Yellow Pages system: first, packages/tools; second, libraries and software.

DS: There is another potential type to be registered, namely 3rd party components (sw, libraries) used by the tools developed by the Centers.

DK: The IATR model has some provisions for registering such 3rd party components.

BL: Notices that recently IATR has been adding some other potentially useful features like a proposal for a controlled vocabulary for the annotation of tools.

KS: Let's go back to the discussion of a charter for the Yellow Pages WG.

Dan Rubin (DR): There are several high level objectives of varying scope that a solution could address: (1) the simplest goal is to just support a simple functionality for free text tool searches, (2) a more sophisticated objective could be to allow browsing/grouping of the NCBC tools based on some more semantically meaningful classification, e.g. using an ontology, (3) a yet wider scope would be to create a more general purpose infrastructure that coulb be potentially open not only to NCBCs but to any researcher who has a tool to register.

Jody Sachs (JS): Representative of NECTAR, another Roadmap initiative that targets the creation of an interoperaility infrastrucure for the exchange of clinical data. Provided overview of NECTAR's mission. Expressed interest in hearing how the NCBCs are planning to address interoperability needs and how these approaches may be applicable in the NECTAR program which is trying to figure out how to navigate the verious ongoing clicnical data interchange standards (efforts like CDISC, HL7, etc).

Mark Musen (MM): There is an apparent abundance of efforts/workshops to create or/and adopt domain ontologies coming from many NIH institutes. Wondered if NIH could provide some way coordinate such activities.

JS: She agrees, they are struggling with the same issue internally at NIH.

PL: Closed the meeting, reminded everyone to register their interest in participating in one of the SDIWG working groups.

Action Items

Prepare and circulate draft charter statement for Yellow Pages working group (Dan Rubin, Ivo Dinov).

SDIWG:Meeting Minutes 20060317

Contents

Preliminary Agenda

Attendees

Minutes

Action Items

Navigation menu

Views

Personal tools

General

Resources

Search

Tools