Skip to main content

Table 3 The LPbyCD performance comparison with a baseline approach and the state of the art

From: LPbyCD: a new scalable and interpretable approach for Link Prediction via Community Detection in bipartite networks

Data set

Approach

Performance

AUC

AUPR

Enzyme

Baseline

0.84

0.15

LPbyCD(Louvain/JC\(_{NC}\))

0.96

0.71

MOLIEREa

0.99

0.90

GPCR

Baseline

0.80

0.22

LPbyCD(Louvain/JC\(_{NC}\))

0.89

0.49

MOLIEREa

0.95

0.75

IC

Baseline

0.76

0.27

LPbyCD(Louvain/JC\(_{NC}\))

0.97

0.78

MOLIEREa

0.98

0.91

NR

Baseline

0.63

0.26

LPbyCD(Louvain/JC\(_{CC}\))

0.82

0.45

MOLIEREa

0.91

0.68

Kinase

Baseline

0.61

0.13

LPbyCD(Louvain/JC\(_{NC}\))

0.91

0.50

ALADINb

0.93

0.60

  1. LPbyCD with the Louvain and JC_{NC} comes close to the performance of the state-of-the-art in terms of AUC (highlighted in bold)
  2. a Buza et al. (2020)
  3. b Buza and Peska (2017)