Skip to main content

Table 3 Power of the UU, UP and REML test procedures

From: Statistical methodology for the analysis of dye-switch microarray experiments

  

μ = 1

μ = 3

Nb Samples

σ B 2 MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacPC6xNi=xH8viVGI8Gi=hEeeu0xXdbba9frFj0xb9qqpG0dXdb9aspeI8k8fiI+fsY=rqGqVepae9pg0db9vqaiVgFr0xfr=xfr=xc9adbaqaaeGacaGaaiaabeqaaeqabiWaaaGcbaacciGae83Wdm3aa0baaSqaaiabdkeacbqaaiabikdaYaaaaaa@2FC9@

UU

UP

REML

UU

UP

REML

5

0.5

5.6

13.6

10.6

55.5

92.1

86.75

5

2

2.8

5.0

12.95

17.9

29.4

34.75

10

0.5

13.2

39.3

33.97

77.8

100.0

99.64

10

2

3.5

7.8

9.06

45.0

63.5

63.06

20

0.5

35.0

80.1

78.13

98.8

100.0

100.0

20

2

7.3

14.5

13.93

82.6

94.8

94.53

30

0.5

51.9

95.5

95.05

100.0

100.0

100.0

30

2

12.1

22.5

21.74

96.2

99.6

99.53

  1. Power (probability of rejecting H0 × 100) of the different test procedures to detect a low (μ = 1, left) or high (μ = 3, right) differential expression.