ABSTRACT
Although carried out with success by hundreds of laboratories, microarray experiments are rather complicated. They involve dozens of steps comprising sample isolation, ampli cation, labeling, hybridization, washing, signal generation, and scanning [1-3]. While, in experienced hands, results are reproducible, precise, and biologically meaningful, the sheer number of steps means that there are multiple places for variability to creep in and cloud biological meaning. This chapter describes a set of experiments that our group routinely performs to assess the performance of microarrays. As array developers, we use these tests when we have created an array of a new design, change manufacturing processes, or modify experimental protocols. These experiments may be of interest to a broader audience of scientists who perform array experiments. This is because the same approaches can be used to qualify the performance of any particular laboratory, laboratory personnel, reagent lots, and so on. In short, whenever a signi cant change is made to the array experimental process, it is good to perform some standard set of tests to verify that performance is maintained and to compare performance over different time periods.
