Figure 2

An example of raw data quality and outlier removal. (panel a, top) Raw data (37 analytes) from array 4 containing all sample replicates shown on an MvA plot (a typical microarray data plot of the log ratio vs. the log difference for each pair of intensities. See: Dudoit, S., Yang, Y. H, Callow, M. J., and Speed, T. P. (2002) Statistica Sinica 12, 111–140). The dashed lines indicate a 99% confidence interval around the data and outliers of this interval are shown in red, black or magenta. (panel b, bottom) Redacted data with 1% of outlier data removed (all points outside of the displayed confidence interval).