eagle-i consortium

pogonotomygobbleΤεχνίτη Νοημοσύνη και Ρομποτική

15 Νοε 2013 (πριν από 3 χρόνια και 6 μήνες)

69 εμφανίσεις

www.eagle
-
i.org

eagle
-
i

making the

invisible visible


Lee M. Nadler, M.D. on
behalf of the eagle
-
I
Consortium

NCRR 56 Day ARRA Challenge

Convene a “diverse” group of at least 6
institutions to deliver:


An approach to identify research resources


A method to catalogue, enter, and store the information
locally


A federated network capable of querying member
institutions and prove that it works


A product that can be validated, exported across America
and sustained


eagle
-
i consortium
--
From Sea to Shining Sea

NINE institutions diverse in geography, culture and resources

Institution

NCRR Programs

Harvard University

CTSA, BIRN, NPRC

Oregon Health &
Science
University

CTSA, NPRC

Dartmouth College

COBRE, INBRE

Jackson State
University

RCMI, RTRN

Montana State
University

COBRE, INBRE

Morehouse School
of Medicine

RCMI, RCRII, RTRN,
CCRE

University of Alaska
Fairbanks

COBRE, INBRE

University of Hawaii
Manoa

RCMI, RCRII, RTRN,
CCRE,
COBRE,
INBRE,

University of Puerto
Rico

RCMI, INBRE,

RTRN, CCHD,

NPRC


Deliver a national research resource discovery network


Onsite teams each capable of discovering and
inventorying research resources


A data inquiry and inventory management system
at each site


Cycles of resource discovery, curation,
dissemination, and assessment


A semantic search application that can find
available research resources that are often invisible


eagle
-
i must create:

Deliverables


Federated system with 9 sites


Effectiveness


“make the invisible visible”


Scalability


Resource types


Quantity of resources


Number of sites


Functionality (obesity use case)

eagle
-
i Architecture


Resource Navigators

Data
Curators


Build Team

eagle
-
i
ontology

Search Application

Federated Network
(SPIN)

Data Entry & Curation Tools

Institutional
Repositories
(RDF)

Data

Key Architecture Elements




Distributed Network


for local control and incremental
expansion


Ontology Driven


for rich search semantics, linking to
outside data and flexibility for change/expansion of resource
types over time


Open Interfaces


for connectivity with outside data and
systems


Data Privacy Controls


to encourage contribution of
“sensitive” resources

Building The Product



Application Team


Data Tools Team


Inventory Management
System Team


Data
Administration


Resource
Navigation

All Sites

Build Team
--

Harvard

Data Curation Teams
(OHSU and Harvard)

Product



Data Models



Ontologies



Inventory Management System



User Interface Query




Research Resources Inventory


Product

Product

Data Curators

Data Entry Tools

Data Tools

Search


Data Entry Tools

Field names and drop
down lists in the data
entry tool are
populated by the
ontology

Finding What You Need

External (Gene/OMIM)

disease

Users may want
to query

eagle
-
i

resource

gene

Users may want
to query

A junior researcher studying obesity wants to
investigate the genetic basis of insulin resistance
in model systems and humans.

Types insulin resistance into the search box

Results are returned for all resources from all institutions related to
insulin resistance.

Interested in reagents thus refines search to reagents only.

The result set was too broad. “Entrez Gene” provides access to genes
related to human disease to help narrow search results.

The investigator wants to find and animal model, so the resource is
refined from insulin resistance to insulin resistance in the mouse.

IRS
-
1 looks promising, so the researcher clicks on the link to go to
Entrez Gene for more information.

The researcher clicks through to Entrez Gene to confirm that IRS
-
1 is a
gene of interest, and searches eagle
-
i for resources related to IRS
-
.1

Plasmids for IRS
-
1 found and the investigator contacts the
researcher to determine their availability.

Much Work Left To Complete During Year 2


Populating resources from all sites, curation, use
cases, sprint test cycles


Improve and expand the system based on user
feedback (integration with PubMed, MGI, other
repositories)


Implement connections to outside systems via
standard interfaces


Begin planning expansion to other institutions



Challenges to Adoption and Sustainability

Develop sustainable models for data collection


Provide value back to the data stewards


Provide value back to the lab

Develop sustainable models for institutional
investment


Ensure that local IT systems are low cost and easy to administer


Provide value back to the institution

Address data privacy concerns


Sensitive resources


Oregon Health and
Science University
(OR)

David W.
Robinson, PhD

University of Alaska
Fairbanks (AK)


Bert Boyer, PhD

University of Hawaii
Manoa (HI)

Richard
Yanagihara, MD

University of Puerto
Rico (PR)

Emma
Fernandez
-
Repollet

Dartmouth College
(NH)

Jason H. Moore,
PhD

Harvard University
(MA)

Lee Nadler, MD;
Douglas
MacFadden

MCS

Jackson State
University (MS)

James L. Perkins,
PhD

Morehouse School of
Medicine (GA)

Gary H.
Gibbons, MD

Montana State
University (MT)

Sara L. Young,
MEd

eagle
-
i consortium