Skip to main content


Figure 4 | BMC Bioinformatics

Figure 4

From: Statistical issues in the analysis of Illumina data

Figure 4

MA-plots comparing the bead summary values for one array with spike concentration 3 pM to an array with spike concentration 1 pM. An increased density of points is indicated by darker shades of blue. Red points highlight the spikes. The horizontal line at M = 1.73 represents the expected log-ratio for the spikes, and the line at M = 0 is the desired level for the remaining non-spikes. Each panel shows the data processed using different background correction methods. Panel A shows the data with no background adjustment, while in panel B local background has been subtracted and in panel C the data has been background subtracted and background normalised. When the data are background subtracted, the range of M and A-values increases and the spikes are closer to the true value than for the non-adjusted data. Background normalistion produces the most variable M-values and over-estimates the M-values for the spikes.

Back to article page