Table 4 The DNN model performance was completely evaluated using different techniques for sequence formulation.

From: Sequence based model using deep neural network and hybrid features for identification of 5-hydroxymethylcytosine modification

Methods

Acc%

Sn%

Sp%

F1 score

MCC

Kmer

65.13

71.34

63.23

52.14

0.303

RC-Kmer

67.34

65.43

70.65

53.78

0.294

PseDNC

68.09

67.32

69.33

55.21

0.353

PseTNC

70.94

80.79

69.84

70.03

0.378

TAC

68.95

79.65

67.83

69.08

0.344

TCC

66.32

69.54

73.43

59.92

0.312

DCC

67.87

74.23

63.67

63.80

0.324

Hybrid feature (without feature selection)

76.87

75.75

84.86

76.29

0.584

Hybrid feature (with feature selection)

84.07

90.29

83.07

84.64

0.736

Back to article page