Figure 5

Agreement of normalization between all data and a subset of data. Normalizations were carried out from perfect-match probes of all probes (blue) or a subset of probes (green) that are spiked in to be unaltered at transcript levels between C and S samples of Golden Spike data. Fitted normalization curves over all (red) or the subset (black) data are shown in MA plots of non-replicate arrays (A) and replicate arrays (B). Microarrays used in A are C2 as baseline array vs. S2. Data used in B are C2 as baseline array vs. C1. There are approximately 34% gene effects in non-replicate arrays. C1, C2: C chip replicate 1, 2, respectively. S2: S chip replicate 2. It should be noted that the two normalization curves generated by Xcorr in A are not significantly different in this non-replicate array pair as defined in Fig. 6.