Annotation-specific expression quotients

Expression quotients were calculated for each 3-level gene ontology term mapped to activation-associated genes. The quotients are calculated as the mean of the absolute log2 ratio of a category divided by the global log2 ratio of the chip. A single asterisk indicates that the 95% confidence interval for the quotient did not include 1. Differences observed in n (number of genes with array data) and N (number of annotated genes in the suppression subtractive hybridization and public EST databases) are due in combination to probe failure as well as an inability to produce unique probes for some genes.