Copyright information:Taken from "Microarray amplification bias: loss of 30% differentially expressed genes due to long probe – poly(A)-tail distances"http://www.biomedcentral.com/1471-2164/8/277BMC Genomics 2007;8():277-277.Published online 15 Aug 2007PMCID:PMC2000903.T2 with wide 95% limits of agreement of -0.7 to 0.7. The variability within the same amplification method is much smaller for as well (1× as () 2x, both with much narrower 95%-limits of agreement of -0.3 to 0.3. No systematic bias was found. As an example, only Cy5 is shown here. T1-1x, tumor – one round of amplification; T1-2x, tumor – two rounds of amplification; T2-1x, duplicate tumor – one round of amplification; T2-2x, duplicate tumor – two rounds of amplification; dotted line, 95%-limits of agreement.