Skip to main content

Table 2 Top ten routes for the Platinum Spike dataset.

From: Preferred analysis methods for Affymetrix GeneChips. II. An expanded, balanced, wholly-defined spike-in dataset

Rank Background correction Probe normalization methods Probe normalization groups PM correction Summarization Probe set normalization methods Probe set normalization groups DEG tests rAUC0.05 TPR0.05 TP0.05*
1 gcrma-reb scaling all pmonly medianpolish vsn technical CyberT 0.848 0.880 1710
2 none vsn all pmonly medianpolish constant all SAMR 0.847 0.883 1717
3 gcrma-reb scaling all pmonly medianpolish vsn technical SAMR 0.847 0.877 1705
4 gcrma-reb constant all pmonly medianpolish vsn technical CyberT 0.846 0.879 1709
5 gcrma-reb scaling all pmonly medianpolish vsn technical LIMMA 0.846 0.878 1707
6 gcrma-reb constant all pmonly medianpolish vsn technical LIMMA 0.845 0.878 1706
7 Alchemy 0.845 0.877 1705
8 none vsn all pmonly medianpolish constant all LIMMA 0.845 0.883 1716
9 none vsn all pmonly medianpolish quantiles technical Fold Change 0.844 0.877 1704
10 rma vsn technical pmonly medianpolish scaling all SAMR 0.844 0.878 1707
  1. *the number of correctly detected DEGs at false positive rate cutoff 0.05.