Skip to main content

Table 2 5-Fold cross-validation performances of methods on Guo dataset

From: xCAPT5: protein–protein interaction prediction using deep and wide multi-kernel pooling convolutional neural networks with protein language model

Method

Accuracy (%)

Precision (%)

Recall (%)

Specificity (%)

F1-Score (%)

MCC (%)

PIPR (2019)

96.47 ± 0.21

96.31 ± 0.23

96.67 ± 0.22

96.65 ± 0.22

96.48 ± 0.20

92.45 ± 0.42

FSNN-LGBM (2021)

98.46 ± 0.20

98.73 ± 0.25

98.18 ± 0.18

98.74 ± 0.25

98.45 ± 0.20

96.92 ± 0.39

MARPPI (2023)

96.03 ± 0.76

98.12 ± 0.98

93.51 ± 1.22

98.82 ± 0.25

NA

91.83 ± 1.32

TAGPPI (2022)

97.81

98.10

98.26

98.10

97.80

95.63

HNSPPI (2023)

98.57 ± 0.11

98.30 ± 0.22

98.85 ± 0.13

NA

98.57 ± 0.11

NA

Our xCAPT5

99.76 ± 0.05

99.76 ± 0.04

99.75 ± 0.07

99.77 ± 0.04

99.37 ± 0.27

99.52 ± 0.10

  1. NA denotes that data is not available. Report with mean and standard deviation. The bold is the best performance in each metric