ABSTRACT
This chapter focuses on the characteristics of useful and meaningful toxicogenomic databases. It discusses the aspects of using toxicogenomics in assembling different types of reference databases and the considerations necessary for employing such databases for drug-development decision making. A database could be queried with novel gene expression profiles from uncharacterized compounds. Samples in this case are messenger RNA from animal tissues or cells that have been exposed to a specific compound or other toxic challenge. Fundamental differences exist between databases for toxicogenomics and sequence assemblies. Toxicogenomics databases consist of sets of empirically derived gene expression profiles. If the intent of the genomics database is to profile large numbers of compounds rapidly, then an in vitro model system is clearly the best choice. In fact, one of the earliest studies in the field of toxicogenomics utilized the in vitro approach to screen compounds and characterize toxicity on the basis of gene expression profiling.
