Figure 1. The flow of work from array interpretation to the generation of a usable database. Each of the three data sets was analyzed with a parsing script developed for each database, dbEST, UniGene, and Gene. The relational database design was dependent on the data we chose to harvest and is described further in Figure 2.