Skip to main content

Table 2 Topological properties of regulated reaction networks and causality graphs built from various experimental lists

From: Using a large-scale knowledge database on reactions and regulations to propose key upstream regulators of various sets of molecules participating in cell metabolism

Hubc

Dataseta

List 1

List 2

List 3

Neighborhoodb

1

2

3

1

2

3

1

2

3

0

Regulated reactions d

         

Molecules, %

1.9

17

52

0.8

10

46

0.1

1.8

14

Reactions, %

1.1

16

59

0.6

9.2

53

0.05

1.6

15

γ e

2.1

2.2

2.3

2.3

2.2

2.3

1.9

2.3

2.2

r

84

93

96

89

90

96

92

94

94

L

3.6

4.2

-

4.4

4.1

-

2.9

48

-

D

6

9

-

8

14

-

4

11

-

Causality graph

         

Nodes, %

1.8

19

61

0.6

10

55

0.09

1.4

16

Edges, %

2.6

28

13

0.1

16

72

0.06

0.3

19

γ

1.69

1.87

1.94

2.35

1.87

1.93

1.67

2.35

1.91

r

79

78

85

91

66

85

83

94

75

100

Regulated reactions d

         

Molecules, %

1.4

11

44

0.8

6.9

35

0.1

1.8

12

Reactions, %

0.8

8

49

0.6

5.7

38

0.05

1.6

11

γ

2.10

2.25

2.30

2.34

2.21

2.30

1.89

2.30

1.24

r

87

97

98

89

96

98

92

94

97

L

3.8

4.5

7.7

4.4

4.4

-

2.9

4.8

4.4

D

6

9

12

8

14

-

4

11

16

Causality graph

         

Nodes, %

1.2

10

50

0.6

6.7

40

0.09

1.4

13

Edges, %

1.3

7.1

41

0.1

3.5

29

0.06

0.3

6.6

γ

1.73

1.97

2.03

2.35

2.06

2.04

2.35

2.09

1.67

r

84

93

91

91

89

91

83

94

93

1000

Regulated reactions d

         

Molecules, %

0.7

3.8

20

0.8

3.6

16

0.1

1.8

7.1

Reactions, %

0.4

1.9

22

0.6

2.4

14

0.05

1.6

5.8

γ

2.15

2.21

2.42

2.34

2.24

2.38

1.89

2.30

2.28

r

94

97

96

89

97

96

92

94

97

L

3.8

4.4

-

4.4

5.0

5.1

2.9

4.8

-

D

6

9

-

8

15

13

11

-

4

Causality graph

         

Nodes, %

0.6

3.0

23

0.6

3.0

15

1.4

6.6

0.09

Edges, %

0.6

2.2

8.9

0.1

1.0

5.3

0.3

2.3

0.06

γ

1.72

1.88

2.34

2.35

2.28

2.03

2.35

2.35

1.67

r

86

87

97

91

97

91

94

97

83

  1. aThree case-study situations were analyzed by using lists with different number and nature of targets: i) the list 1 included 250 unique genes targeted by PPARA[20], ii) the list 2 included 136 gene transcripts that were either up- or down-regulated after addition of agonists of PPARA in cell culture[21], and iii) the list 3 consisted in seven metabolites involved in the successive steps of glycolysis in mammalian cells[32]. Details are provided in Additional file1: Table S1.
  2. bThe first level neighborhood was obtained by taking all the reactions in which the input molecules were involved. The second level was obtained by adding all the new molecules involved in these latter reactions. The third level was an iteration of this procedure.
  3. cThroughout the neighborhood computation, the first n molecules involved in most reactions in cell metabolism were ignored. In the tested situations, n corresponded to 0, 100 or 1,000. After neighborhood computation, these molecules (so-called “hubs” because they shared many relationships) were added to the network only in the case where they participated in reactions selected from the input lists.
  4. dThe proportion (%) of molecules and reactions (nodes and edges, respectively) that were selected from the full graph in the regulated reaction network (causality graph, respectively) was calculated.
  5. eThe topological properties of the network were analyzed using different network statistical parameters. A r value close to 1 indicates that the graph was scale-free. Assuming that the probability P(k) that a molecule in a network interacts with k other molecules follows a power law [P(k) ~ k -γ], a high γ value indicates that there were few highly connected nodes in the network. The quantity L denotes the average shortest path length by which one can reach node A by node B, and D corresponds to the graph diameter. Within a row, the sign “–“ indicates that the computation of these parameters had failed. This analysis shows that the conversion of regulated reactions in a causality graph kept the original structure of the network.
Back to article page

BMC Systems Biology

ISSN: 1752-0509

Contact us