Skip to main content

Table 2 Percentage of time model under H1 is favored over model under H0 for different scenarios For a given scenario, the rows indicate the model under H1 while the columns indicate the model under H0. The (i, j)th element in the matrix represents the percentage of time the model in the i th row is favored over that in the j th column.

From: A new mixture model approach to analyzing allelic-loss data using Bayes factors

Scenario 1 (α0 = 0.22, α1 = 0.66)

H1/H0

2 bin*

2 bb/bin

2 bb

1 bb

1 bin

2 bin

0

21

75

81

100

2 bb/bin

10

0

80

80

100

2 bb

5

0

0

50

98

1 bb

5

0

0

0

100

1 bin

0

0

0

0

0

Scenario 2 (α 0 ~ β (0.26,0.07))

H1/H0

1 bb

2 bin

2 bb/bin

2 bb

1 bin

1 bb

0

22

21

49

75

2 bin

16

0

24

44

72

2 bb/bin

7

14

0

26

74

2 bb

0

12

0

0

68

1 bin

7

0

7

18

0

Scenario 3 (α0 = 0.22, α1 = (1, 0.80, 0.64, 0.43, 0.43))

H1/H0

2 bb/bin

2 bb

2 bin

1 bb

1 bin

2 bb/bin

0

78

79

98

100

2 bb

1

0

31

100

100

2 bin

0

28

0

87

100

1 bb

0

0

5

0

100

1 bin

0

0

0

0

0

Scenario 4 (α0 ~ β (0.26, 0.07), α1 = (1, 0.80, 0.64, 0.43, 0.43))

H1/H0

2 bb

2 bb/bin

1 bb

2 bin

1 bin

2 bb

0

35

75

97

100

2 bb/bin

9

0

54

99

100

1 bb

0

5

0

72

100

2 bin

0

0

9

0

100

1 bin

0

0

0

0

0

  1. *2 bb: Two-component beta-binomial.
  2. 2 bb/bin: Two-component beta-binomial/binomial.
  3. 2 bin: Two-component binomial.
  4. 1 bb: One-component beta-binomial.
  5. 1 bin: One-component binomial.