Skip to main content

Table 1 Sensitivity and complementary specificity for BOLS and SBL algorithms

From: Bayesian Orthogonal Least Squares (BOLS) algorithm for reverse engineering of gene regulatory networks

ε

Sensitivity

C. Specificity

 

BOLS

SBL

BOLS

SBL

N = 20

    

   0.01

0.9676 ± 0.0192

_

0.0009 ± 6.171e-4

_

   0.05

0.9328 ± 0.0215

_

0.0015 ± 6.143e-4

_

   0.10

0.8900 ± 0.0255

_

0.0022 ± 5.095e-4

_

N = 40

    

   0.01

0.9879 ± 0.0089

_

0.0002 ± 1.731e-4

_

   0.05

0.9632 ± 0.0074

_

0.0005 ± 2.146e-4

_

   0.10

0.9371 ± 0.0235

_

0.0010 ± 3.919e-4

_

N = 60

    

   0.01

0.9863 ± 0.0082

_

0.0001 ± 1.053e-4

_

   0.05

0.9720 ± 0.0097

_

0.0004 ± 2.066e-4

_

   0.10

0.9447 ± 0.0127

_

0.0008 ± 2.326e-4

_

N = 80

    

   0.01

0.9872 ± 0.0095

0.9976 ± 0.0039

0.0002 ± 1.693e-4

0.3007 ± 0.0082

   0.05

0.9694 ± 0.0110

0.9896 ± 0.0064

0.0005 ± 2.054e-4

0.3147 ± 0.0072

   0.10

0.9448 ± 0.0201

0.9814 ± 0.0113

0.0008 ± 3.727e-4

0.3270 ± 0.0084

N = 100

    

   0.01

0.9883 ± 0.0084

0.9988 ± 0.0027

0.0002 ± 1.218e-4

0.2953 ± 0.0063

   0.05

0.9694 ± 0.0121

0.9915 ± 0.0075

0.0004 ± 1.883e-4

0.3030 ± 0.0099

   0.10

0.9517 ± 0.0183

0.9843 ± 0.0089

0.0006 ± 2.494e-4

0.3095 ± 0.0075

  1. N is the number of data points, and ε the noise level. mmax is set to 4 and the number of genes K 100.