Skip to main content

Table 1 Performance of GTC and other machine learning models

From: Prediction of plant secondary metabolic pathways using deep transfer learning

Methods

Accuracy (%)

Precision (%)

Recall (%)

F1_score (%)

RF-based

95.66 ± 0.16***

70.95 ± 0.89***

69.38 ± 0.84***

70.16 ± 0.86***

GCN-based

95.94 ± 0.14***

80.81 ± 0.36***

79.47 ± 1.20**

80.14 ± 0.71***

MLGL-MP

96.45 ± 0.15*

84.67 ± 1.36

80.10 ± 0.78**

82.32 ± 0.60*

GCN + CNN

96.05 ± 0.14***

82.26 ± 0.66**

78.67 ± 1.53**

80.41 ± 0.75***

GAT + CNN

96.51 ± 0.27

83.99 ± 1.73

81.74 ± 0.73

82.84 ± 1.10

SuperGAT + CNN

96.63 ± 0.24

84.30 ± 1.46

82.75 ± 0.81

83.52 ± 1.06

GTC

96.75 ± 0.21

85.14 ± 1.42

83.03 ± 0.92

84.06 ± 0.85

  1. The values represent the mean ± standard deviation obtained through fivefold cross-validation. The best value on the metric is highlighted in bold; The symbols *, **, and *** mean that the performance of GTC is significantly better in the t-test at the p-values less than 0.05, 0.01, and 0.001, respectively.
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us