Resources
Case Study: NIH Neuroscience Microarray Consortium
Client
National Institute of Neurological Disease and Stroke, National Institute of Mental Health, Translational Genomics Research Institute, University of California Los Angeles, Duke University
Industry
Life Sciences
Overview
Sophisticated technologies from the Human Genome Project have enabled researchers to understand the basic tenets of human disease, which will eventually lead to cures. While expression microarrays are now a common genome-scale tool used widely in research, the vast amounts of data created by this technology at distant sites are rarely combined – slowing the progress of medical research. Collaboration and consortia, key elements of the National Institutes of Health (NIH) Roadmap, require sharing what and how experiments are conducted and what the results are in a meaningful way to thousands of geographically disparate scientists. The NIH Neuroscience Microarray Consortium was created to extend the services of three Centers of Excellence to all holders of either a NINDS or NIMH grant, enabling them to further their research by working with acknowledged experts in the field.
Problem
The consortia model requires data be collected, managed and distributed consistently across organizations and to the public or risk contributing to the growing silos of data already found on the Internet. Each center had their own infrastructure, methods, tools, and techniques; were geographically spread across the country; were required to serve a national audience of various levels of familiarity with microarray technology; and the consortium needed to be able to report on the progress to different stakeholders at any time. Several applications were available that would address one set of requirements, but none that would address them all simultaneously, particularly across institutions. No solutions mirrored the natural workflow of experiments, enabled cross-organizational collaboration, administration and billing, displayed project progress with a variety of views and produced verifiable MIAME compliant MAGE-ML.
Solution
The problem demanded a custom solution, so 5AM, guided by leading scientists from TGen, UCLA, Duke, NIH Program Administrators and the Consortium’s Advisory Panel, built an enterprise web system tailored to their explicit needs. 5AM elicited and built a generalized workflow to support the business and advanced technology capabilities each center. We improved the ease of enforcing MIAME compliance by verifying compliance as soon as data was entered and at key milestones in the workflow, using the guidelines, exchange standards and software development kits (SDK) provided by the Microarray Gene Expression Data Society (MGED) as significant inputs. 5AM’s expertise implementing IBM’s Rational Unified Process, using the Unified Modeling Language (UML) and building interoperable J2EE/EJB architected web systems empowered us to incrementally and iteratively build the solution. We continue to maintain and evolve the solution, now entering its fourth generation. The system requires only an internet connection and a web browser, the ability to run Java for file transfer, and works on either the Windows or Macintosh platforms. It can be viewed by visiting http://arrayconsortium.tgen.org". The solution, called 5AM’s Microarray Enterprise Manager (MEM) was recognized by IBM Life Sciences as a showcase solution.
Benefits
The application orchestrates and guides researchers to annotate their experiments appropriately, supports inter-lab standard operating procedures, meets evolving industry standards for projects, samples and their derivative microarray data, while granting control to the users over the publication and visibility of their data. This solution facilitates rapid progress in medical research by organizing and exposing the research effort of hundreds of experimenters, detailing the context of those experiments for replication and publication, and ensures they can easily combine and share genomic data. The Consortium expanded to serve researchers across 15 NIH Institutes (from 2) on June 1, 2005, in part because the solution was scalable and strategically fit the NIH Blueprint for Neuroscience. Yale University was added as a Center for Excellence and the investigator pool that the consortium will serve increased to 10,000 users – executed without re-engineering the solution.