Skip to main content

Table 3 Comparison between the results of SAlign (SA) and existing techniques on network pairs which have low percentage of proteins with experimentally resolved structures on the basis of average AFS and average percentage of aligned nodes w.r.t MF and BP

From: SAlign–a structure aware method for global PPI network alignment

Pairs

Evaluation criteria

Alignment algorithms

SA

\(\hbox {SA}^{\mathrm{mc}}\)

HA

MA

IBN

NET

M++

PRO

Mouse fly

AFS\(_{MF}\)

0.50

0.49

0.42

0.36

0.33

0.32

0.37

0.55*

AFS\(_{BP}\)

0.37

0.36

0.31

0.28

0.24

0.23

0.28

0.40*

Nodes\(_{MF}\)

73

72

67

66

58

57

63

61

Nodes\(_{BP}\)

80

80

76

74

58

60

62

56

Mouse worm

AFS\(_{MF}\)

0.56*

0.54

0.49

0.41

0.30

0.29

0.31

0.52

AFS\(_{BP}\)

0.41*

0.40

0.37

0.30

0.25

0.24

0.25

0.39

Nodes\(_{MF}\)

76

74

73

71

62

62

64

64

Nodes\(_{BP}\)

70

68

67

66

70

72

76

73

Avg.

AFS\(_{MF}\)

0.53

0.52

0.46

0.39

0.32

0.31

0.34

0.53

AFS\(_{BP}\)

0.39

0.38

0.34

0.29

0.25

0.24

0.27

0.40

Nodes\(_{MF}\)

74

73

70

69

60

59

64

63

Nodes\(_{BP}\)

75

74

72

70

66

66

69

65

  1. The particular results of the best aligners are differentiated from other aligners by italic text
  2. SAlign performs well in terms of average AFS w.r.t MF and BP and it also outperforms existing aligners in terms of average percentage of align nodes. ‘*’ shows that the results are statistically significant