Skip to main content

Table 2 Performance metrics for the Independent test set-I

From: mLoc-mRNA: predicting multiple sub-cellular localization of mRNAs using random forest algorithm coupled with feature selection via elastic net

Localization

Performance metrics

Sensitivity

Specificity

Accuracy

MCC

F1-measure

aucROC

aucPR

Cytoplasm (300)

70.00

64.49

65.33

25.22

38.09

74.72

37.88

Cytosol (360)

64.17

72.98

71.37

30.35

45.04

76.30

40.51

EPR (170)

94.59

74.71

75.86

35.62

40.93

73.57

17.80

Exosome (140)

66.43

73.50

72.99

22.53

25.92

76.49

34.84

Mitochondrion (76)

82.89

94.72

94.26

54.26

52.73

95.81

75.60

Nucleus (550)

75.09

69.30

70.91

40.21

59.06

78.75

57.96

Pseudopodium (36)

55.56

65.72

65.53

5.99

5.57

69.56

6.55

Posterior (31)

100.00

93.50

93.60

42.97

32.98

98.45

44.12

Ribosome (306)

66.34

74.76

73.45

32.01

43.72

77.91

37.67

  1. Prediction for the test set was made by passing it through all the 9 binary classifiers corresponding to 9 localizations. The accuracies are found to be consistent with that of training set, except aucPR. The low aucPR may be attributed to the highly unbalanced nature of the test dataset, where for a given location sequences of the remaining locations together constituted the negative set
  2. EPR Endoplasmic reticulum
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us