Figure 7 : In the synthetic dataset with differential gene expression (Fig. 6), the unbalance between over- and under-expressed genes was a key factor in the lowered true positive rate and uncontrolled false discovery rate obtained after Median and Quantile normalization.

From: Variation-preserving normalization unveils blind spots in gene expression profiling

Figure 7

Panels show the true positive rate (a) and false discovery rate (b) as a function of the balance of differential gene expression, (, same number of over- and under-expressed genes; , all DEGs over-expressed; , 75% DEGs under-expressed). Each point in both panels represents the results for one treatment compared to the corresponding control, obtained after applying the four normalization methods (same symbols as in Figs 3 and 6; empty black circles, Median normalization; empty red up triangles, Quantile normalization; filled green circles, MedianCD normalization; filled blue up triangles, SVCD normalization). Differential gene expression was analyzed with R/Bioconductor package limma. The dashed horizontal line in (b) indicates the desired bound on the false discovery rate at 0.05.