Database and Software Development

Jackson Laboratory, Bar Harbor ME

Investigators

Linked publications & trials

Abstract

Underlying MGD's biologically-oriented functionality and user interface is a complex information system comprising myriad database structures and software components. The purpose of Core 1 is to maintain this information system while simultaneously developing and deploying new features and data types in support of the aims of this Program Project. At the center of MGD is a Sybase relational database. It is approximately 11 GB in size and contains over 180 tables as well as other components such as views, stored procedures, and triggers. Additionally, over 54 GB of sequence data are stored outside the relational database. We estimate the doubling period of the relational database and external sequence data to be around 18 months. The software of the current MGD system is organized into 133 distinct custom software components comprising over 9,400 source files with over 830,000 lines of code. These components are deployed across 5 main Unix servers;two of these servers support public access with an average over 2 million web hits per month (over 46 hits/minute). In the current granting period through October, 2005 we have had 17 public software and database releases. Highlights of these changes include changing MGD to handle large structured vocabularies like the Mammalian Phenotype (MP) Ontology and the Gene Ontology (GO), revamping mouse phenotypes and disease models, and integrating sequence data and genome coordinates in with other MGD data. We also redesigned the web interface "look and feel" to provide more information in more compact displays, better navigation, and an easy "one-field" search tool.