Show and tell program 04 2014-09-04

1. NIH IC Applications Show & Tell Program
September 4, 2014

Contents Page Show Time
Introduction ......................................................................................................... 8:35
Prevention Abstract Classification Tool (PACT) ................................................ 2
8:45
NIH Library Technology Sandbox ...................................................................... 3 9:05
Low cost display solution for Large screen TVs using Raspberry Pi.................. 4 9:25
CCR Research Exchange (CREx) ........................................................................ 5 9:45
IdeaScale ........................................................................................................... 6 10:05
Amazon Cloud AWS EC2 ................................................................................... 7 10;25
nVision Data Warehouse .................................................................................. 8 10:45
Meeting Registration System - MREGS ............................................................. 9 11:05
Meeting Registration System - MREGS ........................................................... 10 11:25
NLM MeSH on Demand .................................................................................. 11 11:45

2. 1
Introduction
This event is the fourth quarterly Show & Tell event organized by the Enterprise Architecture program of
National Heart, Lung, and Blood Institute’s Office of Management, IT and Applications Center (ITAC). We
are thrilled to offer this forum for application users and technical teams from across the NIH to share
technology that improves how our work is done, and we look forward to transitioning this stewardship
activity to a trans-NIH group for the future with the potential for developing separate scientific,
administrative, and IT oriented demonstration venues.
Still, while every application may not interest everyone in the NIH, the willingness to share both
technologies and the stories behind them is essential for our collective success.

3. 2
App Name: PREVENTION ABSTRACT CLASSIFICATION TOOL (PACT)
Access URL: N/A
Docs URL: (if any) None
IC: OD/ODP
Submitter Name: Patricia L. Mabry, Ph.D.
Email Address: [email protected] Phone [phone]
Functional Area
Where Used:
Extramural How Acquired? Custom Development
App IT Contact: Rich Panzer
Operating Sys: Windows DBMS SQL
Front End: .NET, HTML, CSS
We intend to: Just an FYI / other
FUNCTIONALITY PROVIDED
The Prevention Abstract Classification Tool (PACT) facilitates coding of grant applications used in
portfolio analysis. ODP is working with NIH Office of Portfolio Analysis (OPA) and Center for Information
Technology (CIT) to create a machine-learning based, computerized portfolio analysis tool to
characterize the Prevention Research Portfolio. This requires “gold standards” (a training set) for the
machine to learn from. “Gold standards” are grant abstracts that have been accurately classified
(manually) according to ODP’s Prevention Research Taxonomy. Many thousand gold standards,
generated by manual classification (aka coding )
, are needed for the machine to “learn.” Accurate coding
requires a team approach: each abstract is read and coded by three individuals who then discuss and
reach a “consensus code” that becomes the gold standard. A 23-page protocol dictates the rules for
coding based on the Taxonomy. For a subset of coded abstracts, a second team of three coders recodes
the abstracts following the same procedure. The two consensus codes are reconciled to yield a revised gold standard, if indicated. This double coding provides quality control.
PACT was requisitioned by ODP to facilitate the coding of the large number of gold standards needed.
PACT is a software tool containing a large database of grant abstracts imported from NIH RePORTER.
PACT provides a user interface for coding abstracts in accordance with the Taxonomy and allows coders
to seamlessly access relevant parts of the protocol in real time while coding, enhancing accuracy and reliability. PACT captures and stores data on individual and consensus codes. Because there are multiple
categories within the Taxonomy to be coded, multiple coders, and multiple teams of coders, PACT
calculates 96 inter- rater reliability statistics (Kappas) for each abstract.
A training and testing environment in PACT parallels the “production” environment. Would- be coders
use PACT in training mode to learn the Taxonomy and protocol by coding a large number of abstracts
selected for their pedagogical value. After training is completed, would-be coders code a test set of
abstracts. Kappas are reviewed to ensure that each coder’s test coding is at or above a predetermined
threshold for inter-rater reliability when compared to the correct codes. During production, Kappas are
used to identify abstracts that need coding by a second team, to identify problem coders, and as
indicators of quality control.
ODP and IQ Solutions, Inc. are showing this app to other NIH staff in order to demonstrate the utility of
PACT for supporting abstract coding for portfolio analysis. We think others who are doing any type of
portfolio analysis that relies on manual coding will be interested in our approach and the software that
facilitates this process. Interested parties may contact ODP or IQ Solutions about adapting the approach

4. 3
or tool for their own needs.
Number and Type of Current Users
The current users are ODP staff and staff at IQ Solutions who have been hired by ODP to perform
abstract coding.
E
XTERNAL DATA USED
External data consists of abstracts imported from NIH RePORTER (IMPAC II database).
A
DDITIONAL NOTES
Contact Patricia Mabry, Ph.D. at ODP ([email protected]
) or Rich Panzer at IQ Solutions
([email protected]) for further information about PACT .



App Name: NIH LIBRARY TECHNOLOGY SANDBOX
Access URL: http://nihlibrary.nih.gov/Sandbox
Docs URL: (if any)
IC: NIH Library (ORS)
Submitter Name: Ben Hope
Email Address: [email protected] Phone [phone]
Functional Area
Where Used:
Intramural & training programs How Acquired? Library Sponsored
App IT Contact: Ben Hope (NIH Library)


Operating Sys: n/a DBMS n/a
Front End: n/a
We intend to: Partner on future development or acquisitions
FUNCTIONALITY PROVIDED
The Technology Sandbox is designed to highlight technology-based projects at the NIH, inspire curiosity
and exploration in emerging technologies, and facilitate the development of meaningful partnerships
between the NIH Library staff and its customers.
N
UMBER AND TYPE OF CURRENT USERS
All of NIH and HHS are welcome
A
DDITIONAL NOTES
The Technology Sandbox has a physical presence in the NIH Library in building 10 and a virtual presence
on http://nihlibrary.nih.gov/Sandbox

5. 4

App Name: LOW COST DISPLAY SOLUTION FOR LARGE SCREEN TVS using
Raspberry Pi
Access URL: http://www.raspberrypi.org
Docs URL: (if any)
IC: NIAID, NIH Library (ORS)
Submitter Name: Dawei Lin, Verma Walker, Ben Hope
Email Address: [email protected],
[email protected],
[email protected]
Phone [phone]
Functional Area
Where Used:
Intramural & training programs How Acquired? Custom development
App IT Contact: Dawei Lin (NIAID), Verma Walker and Ben Hope (NIH Library)


Operating Sys: Raspbian DBMS
Front End:
We intend to: Partner on future development or acquisitions
FUNCTIONALITY PROVIDED
This project is a use case of the Technology Sandbox presented in March, 2014 by NIH Library. It uses
Raspberry Pi, a ~$40 computer, to display rotating static images using Quick Image Viewer (QIV,
http://spiegl.de/qiv/) on large screen TVs via HDMI. The project demonstrated that Raspberry Pi is
simple yet powerful enough to support major information display functions using large screen TVs.
N
UMBER AND TYPE OF CURRENT USERS
NIH Library may use it as a pilot to replace laptop computers to drive large screen displays to save cost
and energy use.
A
DDITIONAL NOTES
Some of the work was done by a fifth grader, which indicated that such projects might be suitable for
future outreach and education activities.

6. 5
App Name: CCR RESEARCH EXCHANGE (CREX)
Access URL: https://nci.assaydepot.com
Docs URL: (if any) Information is available at https://nci.assaydepot.com/pages/faq . CREx is
accessible to NIH IRP using active directory credentials.
IC: CCR/NCI
Submitter Name: Mariam Malik
Email Address: [email protected] Phone [phone]
Functional Area
Where Used:
Intramural How Acquired? Custom Development
App IT Contact: Sherman Tang (Assay Depot)
Front End: Web interface – developed using Ruby on Rails
We intend to: Provide this service to others
FUNCTIONALITY PROVIDED
CREx is a “one -stop-shop” for research services available through over 70 CCR/NCI cores and
collaborative resources, as well as over 10,000 commercial vendors. Services can be quickly and easily
identified using service taxonomy or a Google-like search field, and results refined using available filters
to identify vendors with specific certifications or qualifications. The platform also enables
communication with multiple cores and vendors simultaneously, to gather capabilities and quotes,
exchange files and project reports, through the investigator’s private dashboard. In addition, CREx
allows users to review and rate vendors and their services, thus capturing and sharing the organization’s
collective consumer experience and knowledge of best practices.
N
UMBER AND TYPE OF CURRENT USERS
More than 400 NCI scientists (Principal Investigators, postdoctoral fellows, staff scientists etc.) are using
the platform to search and compare research services available through internal NCI cores and external
vendors.
EXTERNAL DATA USED
NCI core managers maintain and update information on the CREx backoffice website, while commercial vendor information is maintained and updated by Assay Depot, and used on their public site
(assaydepot.com) and for all their private research exchanges.
A
DDITIONAL NOTES
Broad adoption throughout the NIH would enhance transparency of services and resources availa ble
through the other IRP cores, and facilitate sharing of core resources amongst the NIH Intramural
Program. In addition, the platform can enable NIH to benefit both from their collective consumer
experience and from savings of research dollars, based on economies of scale or pre-negotiated pricing.

7. 6

App Name: IDEASCALE
Access URL: n/a
Docs URL: (if any) http://www.ideascale.com/
IC: NHLBI
Submitter Name: David Phillips
Email Address: [email protected] Phone [phone]
Functional Area
Where Used:
Administrative How Acquired? COTS/GOTS Procurement
App Acq Contact: Katherine Nguyen (CIT Acquisitions)
Operating Sys: Web- cloud based
DBMS: Cloud
Front End: Major browsers

FUNCTIONALITY PROVIDED
Ideation Programs allow stakeholders to generate, develop, rate, communicate, improve and overall
engage in a process that leads to innovative ideas to address organizational issues and challenges. There
is a growing recognition across many branches of the US Government that better mechanisms are
needed for harnessing the collective talent and expertise of agency employees, as well as external
stakeholders including the general public, to solve problems. Too often, identifying and connecting
innovative thinkers who are most capable of generating truly pioneering solutions can be challenging.
IdeaScale is a Cloud-based platform which NHLBI OM ITAC is using for its regular IT Needs Data Call to
users across the IC and which our IC Director intends to leverage for a strategic visioning process.

NUMBER AND TYPE OF CURRENT USERS
Several hundred, across all functional areas.

EXTERNAL DATA USED
None

8. 7

App Name: AMAZON CLOUD AWS EC2
Access URL: http://pub.ncibbrb.org/ , others
Docs URL: (if any) http://aws.amazon.com/
IC: NCI
Submitter Name: David Tabor
Email Address: [email protected] Phone [phone]
Functional Area
Where Used:
IT How Acquired? COTS/GOTS Procurement
App IT Contact: David Tabor ([email protected], [phone]
Operating Sys: Windows / linux DBMS Any, n/a
Front End: Web-based
We intend to: Just an FYI / other
FUNCTIONALITY PROVIDED
We have found that using the Amazon Cloud as an adjunct to our in- house Data Center allows us to very
rapidly configure and deploy new applications, demo and train for internal and external customers,
stand up and configure servers and work out the “kinks.” It also provides a platform for secure
collaboration inside and outside the enterprise, and a unique platform for monitoring our public-facing
application, http://biospecimens.cancer.gov/brd/

NUMBER AND TYPE OF CURRENT USERS
The version of Biospecimen Research Database ( BRD) we have running in the Amazon C loud is for User
Acceptance Testing (UAT). We also have a Windows server that we use to test access to BRD through
the firewall.
The BRD is a free and publicly accessible literature database that contains peer- reviewed primary and review
articles in the field of human Biospecimen Science. Each entry has been created by a Ph.D. level scientist to
capture
(1) relevant parameters that include the biospecimen investigated (type and location, patient diagnosis),
preservation method, analyte(s) of interest and technology platform(s) used for analysis;

(2) the pre-analytical factors investigated, including those encountered during the lifecycle of a biospecimen
(e.g. ischemia time, fixation parameters, storage conditions); and

(3) an original summary of relevant results.
The Comprehensive Data Resource (CDR) is a distributed web- based bioinformatics system that
manages and maintains multi-dimensional data models on biospecimens. The CDR was developed and is
currently utilized to collect data on biospecimens obtained from cancer patients and post- mortem
donors, for the Biospecimen Pre-analytical Variables (BPV) and Genotype-tissue Expression (GTEx)
programs. The CDR is an extranet hosted at the NCI Frederick. We use the Amazon Cloud to host a
training server for the CDR, with test data, that does not have to be behind the NIH firewall.
Grand total over 150 users.
E
XTERNAL DATA USED
PubMed, numerous scientific journals, UMLS

9. 8

App Name: NVISION DATA WAREHOUSE
Access URL: https://nvision.nih.gov/nVision_Portlets/nVision_Launch_Pad/nVision_User_Acces
s.cfm
Docs URL: (if any) https://nvision.nih.gov/
IC: CIT-Center for Information Technology
Submitter Name: Denise Dmuchowski
Email Address: [email protected] Phone [phone]
Functional Area
Where Used:
Administrative How Acquired? Custom Development
App IT Contact: [email protected]
Operating Sys: Mix of Microsoft Windows and
Unix
DBMS Oracle 11g
Front End: Mix of Adobe ColdFusion Web Apps, SAP BusinessObjects, and QlikView Applications (I’m leaving our Hummingbirsd and IBM Cognos)
We intend to: Provide this service to others
FUNCTIONALITY PROVIDED
nVision is the enterprise information solution that delivers a centralized data repository, desktop
analytics and reporting tools to streamline business management at the NIH. nVision gathers data from XX source systems and then applies business rules to over 59,000 business transactions per day. The
result is useful, actionable information placed at the fingertips of 4,000 users and available to over 100 key business systems across the NIH. nVision provides information necessary to achieve essential business objectives via standard and custom reports, dashboards, and analytics.
N
UMBER AND TYPE OF CURRENT USERS
Administrative management community ~4,000 users
E
XTERNAL DATA USED
25 systems feed nVision data

10. 9
App Name: MEETING REGISTRATION SYSTEM - MREGS
Access URL: Sample site: http://meetings.nigms.nih.gov/meetings/NIGMSWebaaps/
Docs URL: (if any) n/a
IC: NIGMS
Submitter Name: Alex Naneyshvili
Email Address: [email protected] Phone [phone]
Functional Area
Where Used:
Administrative How Acquired? Custom Development
App IT Contact: Anjum Dahya 301-435-8575
Operating Sys: Windows 2003 DBMS Oracle 11gR2
Front End: ColdFusion 8
We intend to: Provide this service to others
FUNCTIONALITY PROVIDED
The NIGMS Meeting Registration System (MREGS) is a web -based tool for organizers of NIGMS-
sponsored conferences and meetings held on or off the NIH campus.
Features of MREGS:
 Publishes meeting information including agenda and logistics
 Provides web-based meeting registration
 Allows the meeting organizer to:
 manage registration process
 generate customized rosters
 send e-mails to registrants
N
UMBER AND TYPE OF CURRENT USERS
MREGS is currently used by NIGMS, FIC, NINR, OD, and NINDS
E
XTERNAL DATA USED
Meeting Registrants’ Data only.

11. 10
App Name: MEETING REGISTRATION SYSTEM - MREGS
Access URL: Sample site: http://meetings.nigms.nih.gov/meetings/NIGMSWebaaps/
Docs URL: (if any) n/a
IC: NIGMS
Submitter Name: Alex Naneyshvili
Email Address: [email protected] Phone [phone]
Functional Area
Where Used:
Administrative How Acquired? Custom Development
App IT Contact: Anjum Dahya 301-435-8575
Operating Sys: Windows 2003 DBMS Oracle 11gR2
Front End: ColdFusion 8
We intend to: Provide this service to others
FUNCTIONALITY PROVIDED
The NIGMS Meeting Registration System (MREGS) is a web-based tool for organizers of NIGMS-sponsored conferences and
meetings held on or off the NIH campus.
Features of MREGS:
 Publishes meeting information including agenda and logistics
 Provides web-based meeting registration
 Allows the meeting organizer to:
 manage registration process
 generate customized rosters
 send e-mails to registrants
NUMBER AND TYPE OF CURRENT USERS
MREGS is currently used by NIGMS, FIC, NINR, OD, and NINDS
E
XTERNAL DATA USED
Meeting Registrants’ Data only.
A
DDITIONAL NOTES
[Type any additional notes if needed.]

12. 11
App Name: NLM MESH ON DEMAND
Access URL: http://www.nlm.nih.gov/mesh/MeSHonDemand.html
Docs URL: (if any) n/a
IC: NLM
Submitter Name: Dan Cho
Email Address: [email protected] Phone [phone]
Functional Area
Where Used:
Intramural How Acquired? Custom Development
App IT Contact: Jim Mork
Operating Sys: Windows 2003 DBMS Oracle 11gR2
Front End: ColdFusion 8
We intend to: Provide this service to others
FUNCTIONALITY PROVIDED
Identifies MeSH Terms in your text using the NLM Medical Text Indexer (MTI) program. After
processing, MeSH on Demand returns a list of MeSH Terms relevant to your text and PMID of top ten
related articles.
N
UMBER AND TYPE OF CURRENT USERS
Peak users at 150,000+ in first month after rollout
E
XTERNAL DATA USED
Meeting Registrants’ Data only.
A
DDITIONAL NOTES
For more information about MeSH on Demand, please see our NLM T echnical Bulletin article and
more
detail here.