Resources
Case Study: UCLA Web Services Customization
Client
University of California Los Angeles
Industry
Life Sciences
Overview
UCLA’s DNA Microarray Core is one of four centers working under the NIH Neuroscience Microarray Consortium and is nationally recognized as a Center for Excellence. The investigators that use their Core Facility come from a variety of places (primarily UCLA researchers) and UCLA created a web system prior to the consortium’s existence for managing projects. UCLA’s Gene Expression Core (GEC), built on the open source ArsDigita Community System (ACS) software, needed to be able to exchange data with the Consortium system, 5AM’s Microarray Enterprise Manager (MEM) – a J2EE/EJB solution built with IBM software (WebSphere and DB2).
Problem
All projects that were assigned to UCLA from the Consortium required double entry on the part of the lab researchers and their customers because regardless of where a project came from, they needed to be tracked in the custom UCLA system. While similar in intent – project tracking, data upload and publication – the two systems had different requirements and design – particularly with regard to the requirement of the evolving microarray standard MIAME (the minimum information about a microarray experiment). In addition, the interactions between the systems would need to be seamless to the internal and external users. Finally, the development effort would need to take place in a nationally distributed environment (from Boston to Phoenix to Los Angeles).
Solution
A collaborative effort was undertaken by 5AM and the UCLA Control team to identify the appropriate steps in the workflow where bi-directional data exchange should occur. After a brief Inception, the two main points were project approval and completion where large amounts of data (100’s of megabytes) would need to acquired, extracted and readied for publication and web services would be the mechanism for data exchange between the systems. The Consortium solution would indicate when a project had been directed to UCLA and pass all data – including the creation of users – to the GEC system. Lab technicians would then use the GEC system to move the project forward. Once the experiment was complete, all project annotation and raw data would be sent from the GEC to the Consortium without any human intervention required. The annotation data would populate appropriate to indicate what work was done, and the raw data was extracted to be readied for analysis and publication to the National Center for Bioinformatics (NCBI)’s Gene Expression Omnibus (GEO) – the largest repository of gene expression data in the world.
Benefits
The most significant benefits were to the end users of both systems, where the amount of work on the part of lab technicians was shortened, double entry was removed and customers could view their project progress and respond to information and billing requests as directed by each system. In addition, the cost was lower then either group performing all the work themselves. The effort (requirements, development, testing) was shared between 5AM and UCLA resources. The final benefit was to the consortium itself, as it now runs even smoother as the center (UCLA) that had a pre-existing web system is now fully integrated into the overall workflow. With the consortium going NIH-wide in June 2005 – moving from supporting users with grants from two Institutes (NINDS & NIMH) to fifteen (15) and roughly 10,000 potential users, the need for efficiency has never been greater.