Additional data file 1.
Eight arbitrarily chosen cancer data sets [8-12] were tested for intensity-dependent correlation bias. For each data set, the expression values were calculated from the raw data using the RMA, MAS5, MBEI, and GCRMA normalization algorithms. Regardless of the normalization algorithm, intensity-dependent correlation bias was present, and PSQN reduced this bias.
Format: PDF Size: 778KB Download file
This file can be viewed with: Adobe Acrobat Reader
Eklund and Szallasi Genome Biology 2008 9:R26 doi:10.1186/gb-2008-9-2-r26