Skip to main content

Table 3 Results obtained using the graphical lasso algorithm.

From: Inference of radio-responsive gene regulatory networks using the graphical lasso algorithm

- lo g 10 ( λ )

TP

FN

FP

# of edges

Precision

Recall

f-score

p- value

0.80

4

24

4

8

0.50

0.14

0.22

0.055

0.96

5

23

7

12

0.42

0.18

0.25

0.062

1.05

8

20

11

19

0.42

0.29

0.34

0.017

1.10

10

18

15

25

0.40

0.36

0.38

0.007

1.20

11

17

16

27

0.41

0.39

0.40

0.005

1.31

12

16

20

32

0.38

0.43

0.40

0.010

1.48

15

13

23

38

0.39

0.54

0.45

0.001

1.52

16

12

25

41

0.39

0.57

0.46

0.001

1.72

17

11

32

49

0.35

0.61

0.44

0.002

1.78

17

11

36

53

0.32

0.61

0.42

0.007

(A) Comparison of the gold-standard network and networks estimated from GSE1977,

- lo g 10 ( λ )

TP

FN

FP

# of edges

Precision

Recall

f -score

p- value

0.55

1

27

5

6

0.17

0.04

0.06

0.737

0.58

2

26

8

10

0.20

0.07

0.11

0.596

0.71

3

25

12

15

0.20

0.11

0.14

0.599

0.80

5

23

17

22

0.23

0.18

0.20

0.420

1.04

7

21

23

30

0.23

0.25

0.24

0.337

1.13

9

19

26

35

0.26

0.32

0.29

0.194

1.15

10

18

27

37

0.27

0.36

0.31

0.150

1.31

12

16

32

44

0.27

0.43

0.33

0.110

1.39

13

15

35

48

0.27

0.46

0.34

0.096

1.46

13

15

38

51

0.25

0.46

0.33

0.135

(B) comparison of the gold-standard network and networks estimated from GSE23393, and

- lo g 10 ( λ ) in GSE1977

- lo g 10 ( λ ) in GSE23393

# of edges in GSE1977

# of edges in GSE23393

# of common edges

f -score

p-value

0.80

0.55

8

6

1

0.14

0.324

0.96

0.58

12

10

2

0.18

0.228

1.05

0.71

19

15

4

0.24

0.123

1.10

0.80

25

22

8

0.34

0.022

1.20

1.04

27

30

12

0.42

0.004

1.31

1.13

32

35

17

0.51

< 0.001

1.48

1.15

38

37

19

0.51

< 0.001

1.52

1.31

41

44

22

0.52

0.001

1.72

1.39

49

48

26

0.54

0.001

1.78

1.46

53

51

29

0.56

<0.001

(C) comparison of networks estimated from GSE1977 and GSE23393.