Skip to main content

Table 4 Differences in classification performance (GEOlimma AUC–Limma AUC) for 10 data subsets of the Asthma comparison. Italic p values (Wilcoxon signed-rank test) denote statistically significant AUC improvements of GEOlimma over Limma

From: GEOlimma: differential expression analysis and feature selection using pre-existing microarray data

Sample order

AUCdiff

Wilcox p value

VadAUCdiff

Wilcox p value

1

0.0075

2.36E−01

\(-\) 0.01425

9.78E−01

2

\(-\) 0.0275

8.53E−01

0.01375

1.43E−01

3

\(-\) 0.0075

6.07E−01

\(-\) 0.025

9.97E−01

4

\(-\) 0.0075

7.79E−01

0.00525

2.39E−01

5

\(-\) 0.0175

9.31E−01

0.0105

3.12E−01

6

\(-\) 0.03

8.97E-01

\(-\) 0.0085

7.23E−01

7

0.0675

3.98E−02

0.06025

9.77E−04

8

0.025

1.04E−01

\(-\) 0.007

7.54E−01

9

0.0525

1.23E−02

0.0385

1.95E−03

10

\(-\) 0.0075

7.36E−01

\(-\) 0.001

7.93E−01

  1. *Asthma dataset on LR classification