Skip to main content

Table 1 Performance evaluation and comparison of SITPR, PCC, MI, CLR, ARACNE, GENIE3 and TIGRESS for networks of three different sizes

From: Systematic identification of transcriptional and post-transcriptional regulations in human respiratory epithelial cells during influenza A virus infection

Method

Node Size

SN

SP

ACC

F-measure

MCC

AUC

SITPR

10

0.886 ± 0.090

0.971 ± 0.069

0.943 ± 0.667

0.925 ± 0.070

0.875 ± 0.144

0.966 ± 0.069

50

0.775 ± 0.089

0.635 ± 0.088

0.670 ± 0.054

0.690 ± 0.039

0.358 ± 0.062

0.703 ± 0.037

100

0.690 ± 0.080

0.478 ± 0.101

0.530 ± 0.079

0.557 ± 0.080

0.147 ± 0.110

0.573 ± 0.071

PCC

10

0.400 ± 0.241

0.486 ± 0.125

0.457 ± 0.133

0.398 ± 0.204

-0.112 ± 0.293

0.596 ± 0.082

50

0.495 ± 0.157

0.453 ± 0.167

0.463 ± 0.124

0.444 ± 0.105

-0.043 ± 0.189

0.557 ± 0.038

100

0.534 ± 0.084

0.526 ± 0.057

0.528 ± 0.045

0.525 ± 0.050

0.052 ± 0.045

0.517 ± 0.027

MI

10

0.243 ± 0.243

0.679 ± 0.097

0.533 ± 0.092

0.293 ± 0.256

-0.101 ± 0.255

0.567 ± 0.062

50

0.590 ± 0.145

0.527 ± 0.106

0.543 ± 0.065

0.542 ± 0.036

0.103 ± 0.072

0.574 ± 0.027

100

0.539 ± 0.063

0.569 ± 0.079

0.561 ± 0.050

0.547 ± 0.030

0.095 ± 0.052

0.568 ± 0.024

CLR

10

0.400 ± 0.148

0.543 ± 0.096

0.495 ± 0.068

0.441 ± 0.081

-0.055 ± 0.143

0.540 ± 0.059

50

0.565 ± 0.133

0.510 ± 0.090

0.523 ± 0.061

0.520 ± 0.072

0.064 ± 0.107

0.539 ± 0.066

100

0.568 ± 0.130

0.543 ± 0.102

0.549 ± 0.059

0.538 ± 0.044

0.098 ± 0.084

0.550 ± 0.039

ARACNE

10

0.500 ± 0.168

0.643 ± 0.181

0.595 ± 0.117

0.532 ± 0.105

0.151 ± 0.213

0.577 ± 0.084

50

0.620 ± 0.118

0.508 ± 0.119

0.535 ± 0.067

0.539 ± 0.0275

0.114 ± 0.062

0.547 ± 0.043

100

0.576 ± 0.112

0.525 ± 0.108

0.537 ± 0.062

0.531 ± 0.039

0.089 ± 0.062

0.554 ± 0.050

GENIE3

10

0.429 ± 0.252

0.521 ± 0.147

0.490 ± 0.163

0.443 ± 0.21

-0.050 ± 0.345

0.609 ± 0.057

50

0.565 ± 0.116

0.444 ± 0.228

0.473 ± 0.157

0.444 ± 0.162

-0.007 ± 0.187

0.529 ± 0.041

100

0.512 ± 0.093

0.544 ± 0.062

0.536 ± 0.043

0.520 ± 0.050

0.049 ± 0.077

0.521 ± 0.033

TIGRESS

10

0.629 ± 0.168

0.586 ± 0.171

0.600 ± 0.127

0.583 ± 0.135

0.205 ± 0.233

0.586 ± 0.122

50

0.490 ± 0.152

0.463 ± 0.105

0.470 ± 0.09

0.461 ± 0.095

-0.041 ± 0.167

0.536 ± 0.067

 

100

0.554 ± 0.127

0.507 ± 0.083

0.519 ± 0.058

0.515 ± 0.070

0.053 ± 0.105

0.549 ± 0.033

  1. For fairness of comparison when evaluating the performances, all the methods are assigned the same task (that is, identify the activated regulatory relationships for a given background network).