Statistical design of reverse dye microarrays.

摘要: Motivation: In cDNA microarray experiments all samples are labelled with either Cy3 dye or Cy5 dye. Certain genes exhibit bias—a tendency to bind more efficiently one of the dyes. The common reference design avoids problem bias by running arrays ‘forward’, so that being compared always same But comparison different dyes is sometimes interest. these situations, it necessary run some ‘reverse’—with labelling reversed—in order correct for bias. will impact one’s ability identify differentially expressed in tissues conditions. We address issue how many specimens needed, forward and reverse perform, optimally assign labels specimens. Results: consider three types which needed: paired samples, from two predefined groups, data when present simple probability models data, derive optimal estimators relative gene expression, compare efficiency a range designs. each case, we sample size formulas. show individual generally not required.