Table 2 Rank distributions denoting performance based on statistical feature selection and batch effect correction using D2.2 data (c.f. Fig. 4B).

From: How to do quantile normalization correctly for gene expression data analyses

Batch F-score Delta
0 0.2 0.5 0.8 0 0.2 0.5 0.8
CEP: 0
Adjust 0 0 0 0 1 4 4 4
Class-specific 0 0 0 0 2 4 4 4
qsmooth 0 0 0 0 4 4 4 4
All 0 0 0 0 4 4 4 4
Discrete 0 0 0 0 4 4 4 4
Ratio 0 0 0 0 6 1 1 1
CEP: 0.2
Adjust 2.5 2.5 2 4 2.5 4 4.5 4.5
Class-specific 2.5 2.5 2 1 1 1 1 1
qsmooth 2.5 2.5 5 4 4 2 2 2
All 5.5 6 5 4 5 4 3 3
Discrete 2.5 2.5 2 4 2.5 4 4.5 4.5
Ratio 5.5 5 5 4 6 6 6 6
CEP: 0.5
Adjust 3 3 3 4 2.5 3.5 4.5 4.5
Class-specific 3 3 3 2 2.5 1 1 1
qsmooth 3 5 5 4 2.5 2 2 2
All 6 6 6 6 5 5 3 3
Discrete 3 3 3 4 2.5 3.5 4.5 4.5
Ratio 3 1 1 1 6 6 6 6
CEP: 0.8
Adjust 3.5 4 4 3 2.5 3.5 3.5 4.5
Class-specific 3.5 4 2 3 2.5 1 1 1
qsmooth 3.5 4 4 5 2.5 2 2 2
All 6 6 6 6 5 5 5 3
Discrete 3.5 2 4 3 2.5 3.5 3.5 4.5
Ratio 1 1 1 1 6 6 6 6
  1. The top ranked method is displayed in bold.