Skip to main content
Figure 3 | BMC Bioinformatics

Figure 3

From: A power law global error model for the identification of differentially expressed genes in microarray data

Figure 3

Performance of PLGEM-STN in ranking differentially expressed genes. ROC plots were used to compare the sensitivity vs. specificity trade-off of the following three statistics: PLGEM-STN (black), CLASSIC-STN (blue) and SAM-STAT (red). SAM-STAT values were obtained using the R package "siggenes" [21]. Absolute values of the corresponding statistics were sorted in decreasing order, first n genes were selected (where n ranged from 5 to 200) and false positive and false negative rates were evaluated on the HG-U133A Latin Square dataset. Note that, while the transcripts in Exp02 (Exp14) are spiked-in at twice (half) the concentration than in Exp01, in both Exp06 vs. Exp01 and in Exp10 vs. Exp01 comparisons the nominal fold-change of spiked-in transcripts ranged from 32 to 512.

Back to article page