DEVELOPMENT OF A RELATIONAL TOXICOGENOMICS DATABASE FOR THE PREDICTION OF CHEMICAL TOXICITY.

P C Boutros1,2,3, L D Burgoon2,3,4 and T R Zacharewski1,2,3. 1Dept. of Biochemistry & Molecular Biology, 2Institute for Environmental Toxicology, 3National Food Safety & Toxicology Center, 4Dept. of Pharmacology & Toxicology, Michigan State University, East Lansing, MI.

Toxicogenomic studies generate vast amounts of data that require proper storage in order to facilitate data handling and analysis. The inability to relate changes in global gene expression to specific genes, their function, and other information further complicates data analysis and interpretation. To automate these essential functions, dbZach, a database operating under the Oracle relational database management system has been built. dbZach provides a core database that handles all aspects of our multi-species toxicogenomic/microarray investigations including: 1) storage of important cDNA/EST clone information, such as sequence, 2) frequent, automated updates of gene annotation information, 3) microarray data storage, 4) physical mapping data for each microarray utilized and constructed within the lab. Currently, dbZach is comprised of four main subsystems: 1) the Clones Subsystem, 2) the Microarray Subsystem, 3) the Gene Function Subsystem, and 4) the Lookup Subsystem. The subsystems interrelate to find and filter data within a dataset, and define relationships between tables. Querying dbZach for information regarding mTZ2.0, our mouse microarray, reveals the availability of 2,288 cDNAs/ESTs representing 1,948 unique genes. Ninety-five percent of the available cDNAs/ESTs can be assigned to a specific cluster with the average cluster represented by 1.25 clones. Less than 1% of the genes on our murine microarray, mTZ2.0, are represented by more than 3 clones. As part of our effort to develop a transcript profile database predictive of chemical toxicity, future planned developments for dbZach include compliance with the MIAME standards for microarray experiments, statistical analysis functionality, and comparisons of global gene expression data stored within dbZach.