Figure 5
Chip-type and RNA effect on the sensitivity profiles of rank r = 1 - 4. The profiles refer to different hybridizations which use either identical mRNA (human reference RNA) and different chip types (panel a, HG U133A versus HG U133plus2) or arrays of the same type (MG-230 2.0) but different RNA samples (panel b). The profiles of the 'HG133plus2_mask' analysis (part a) refer to the same collection of probe sequences as used for calculating the HG133_Z profiles. In general, the sensitivity profiles are relatively invariant in the different situations except the amplitude of the (GGG)1-effect which considerably scatters in different hybridizations to the same chip-type (see part b). Part c of the figure shows the distribution of the relative (GGG)1-amplitude Δ σ (GGG) for 29 independent hybridizations to MG-230 2.0 arrays (the list of GEO-accession numbers is given in [Additional file 1]). Δσ (GGG) is calculated as the difference between the sensitivity values of (GGG)1- and (CCC)12-motifs (see part b for illustration).