highly significantP < 0.01
A 1.5-fold change expression cut-off for ratio experiments was applied together with anticorrelation of dye-swapped ratio profiles, rendering the microarray analysis highly significant ( P < 0.01), robust, and reproducible ( 48 ).