Microarray Normalization Revisited for Reproducible Breast Cancer Biomarkers

Figure 6

Comparison between the 2 top pipelines, fRMA, and GCRMA. For reference, we also show the “dinosaur”, “RMA R single”. The derby “fRMA vs. GCRMA R global” is labelled comparison “K” in Table 2 to Table S1. Comparison of fRMA with the dinosaur “RMA R single” is labelled “M”. “GCRMA R global” versus “RMA R Single” is labelled “N”. (a) Distribution profiles of log₂-RMA normalized expression values. (b) Histogram of correlation distances of log₂-normalized expression values between pairs of pipelines. (c) For all 54675 probe sets of each sample, the 80% quantile (q_0.80) is computed and plotted over sample-ID (-axis).

(a)

(b)

(c)