Skip to main content

Table 5 Predicting performance in identifying eight main NR families based on Feature set 3 and Feature set 5

From: Accurate prediction of nuclear receptors with conjoint triad feature

NR Subfamily

CTF

CTF + AAC

 

Sens(i)

Spec(i)

Acc(i)

MCC(i)

Sens(i)

Spec(i)

Acc(i)

MCC(i)

NR1

158/162 = 0.9753

0.9519

0.9599

0.9135

158/162 = 0.9753

0.9551

0.9620

0.8966

NR2

132/140 = 0.9429

0.9700

0.9620

0.9091

133/140 = 0.95

0.9731

0.9663

0.9189

NR3

77/82 = 0.9390

0.9949

0.9852

0.9479

78/82 = 0.9512

0.9949

0.9873

0.955

NR4

20/23 = 0.8696

1

0.9937

0.9294

20/23 = 0.8696

1

0.9937

0.9294

NR5

27/29 = 0.9310

1

0.9958

0.9627

27/29 = 0.9310

1

0.9958

0.9627

NR6

5/7 = 0.7143

1

0.9958

0.8434

5/7 = 0.7143

1

0.9958

0.8433

NR7

20/21 = 0.9524

1

0.9979

0.9748

20/21 = 0.9524

1

0.9979

0.9748

NR8

8/10 = 0.8

1

0.9958

0.8925

8/10 = 0.8

1

0.9958

0.8340

Overall

447/474 = 0.9430

0.9919

0.9858

0.9349

449/474 = 0.9473

0.9925

0.9868

0.9397

  1. (10-fold cross-validation test)