HowTo plw

This vignette describes how to use plw, an R implementation of the Probe level Locally moderated Weighted median-t (PLW) method (Åstrand et al., 2007a) for finding differentially expressed genes. PLW uses an empirical Bayes model taking into account the dependency between variability and intensity-level. A global covariance matrix is also used allowing for differing variances between arrays as well as array-to-array correlations, and thus PLW performs weighted analysis. PLW is specially designed for Affymetrix type arrays (or other multiple-probe arrays). Instead of making inference on probe-set summaries, comparisons are made separately for each perfect-match probe and are then summarized into one score for the probe-set. The Locally Moderated Weighted-t (LMW) method, applying the model of PLW on probe-set summaries or data from single probe arrays, is also implemented in the plw package. See Åstrand et al. (2007a) for details on PLW and LMW, and Kristiansson et al. (2005, 2006), Sjögren et al. (2007), and Åstrand et al. (2007b) for details on weighted analysis for microarrays. PLW is demonstrated in Sections 2 to 4, and LMW in Section 5.