Skip to main content

Table 3 Comparison between SCADDx and LOADDx using CTD KB considering best parameter values (P & Q) for all the four datasets

From: Enabling personalised disease diagnosis by combining a patient’s time-specific gene expression profile with a biomedical knowledge base

Datasets

Algorithm

Parameter

values

Accuracy

n@1 (%)

Accuracy

n@2 (%)

Accuracy

n@3 (%)

Accuracy

n@4 (%)

Accuracy

n@5 (%)

Accuracy

n@10 (%)

Dataset 1

Testset 1a

(GSE73072)

SCADDx

P = 100

Q = 175

76.92

76.92

76.92

76.92

76.92

84.62

LOADDx

P = 25

Q = 225

69.23

69.23

69.23

69.23

69.23

69.23

Dataset 1

Testset 1b

(GSE73072)

SCADDx

P = 100

Q = 175

84.62

76.92

76.92

76.92

76.92

76.92

LOADDx

P = 25

Q = 225

84.62

84.62

84.62

84.62

84.62

84.62

Dataset 2

Testset 2a

(GSE68310)

SCADDx

P = 150

Q = 300

100

100

100

100

100

100

LOADDx

P = 50

Q = 300

75

81.25

87.5

87.5

93.75

100

Dataset 2

Testset 2b

(GSE68310)

SCADDx

P = 150

Q = 300

86.66

93.33

93.33

93.33

93.33

100

LOADDx

P = 50

Q = 300

80

80

80

80

80

86.66

Dataset 3

(GSE90732)

SCADDx

P = 25

Q = 25

100

100

100

100

100

100

LOADDx

P = 25

Q = 25

75

100

100

100

100

100

Dataset 4

(GSE61754)

SCADDx

P = 25

Q = 25

85.71

85.71

85.71

85.71

85.71

85.71

LOADDx

P = 25

Q = 50

85.71

85.71

85.71

85.71

85.71

85.71