Bioinformatics researchers have learned many things from the analysis of microarray data. For instance, power to detect differential expression can be improved and false discoveries reduced by sharing information across all probes. One such procedure is limma (Smyth, 2004), where an empirical Bayes model is used to moderate the probe-wise variances. The moderated variances replace the probe-wise variances in the t- and F-statistic calculations. In a closely analogous but mathematically more complex procedure, edgeR models count data using an overdispersed Poisson model, and uses an empirical Bayes procedure to moderate the degree of overdispersion across genes.
