Skip to main content

Table 2 The proportion of biomarkers outperforming the AURKA single-gene biomarker on each of the twelve meta-dataset partitions by model

From: A comparative study of survival models for breast cancer prognostication revisited: the benefits of multi-gene models

Partition

single-gene

pairDifference

geneSIMMS-5

geneSIMMS-50

geneSIMMS-100

P1

3.75×10−4

2.25×10−5

1.22×10−3

0.0289

0.070

P2

7.50×10−4

1.75×10−4

6.96×10−3

0.268

0.570

P3

1.13×10−3

9.18×10−5

4.60×10−3

0.208

0.488

P4

6.50×10−3

9.83×10−4

0.0178

0.380

0.635

P5

8.75×10−4

2.44×10−5

6.36×10−4

0.129

0.334

P6

3.13×10−3

1.43×10−3

0.0217

0.407

0.624

P7

7.50×10−4

1.31×10−6

1.15×10−4

7.72×10−3

0.0292

P8

1.38×10−3

5.69×10−3

0.0721

0.925

0.992

P9

1.25×10−4

3.72×10−5

1.03×10−3

0.199

0.483

P10

0

0

3.00×10−6

1.23×10−4

4.27×10−4

P11

6.25×10−4

2.11×10−4

5.39×10−3

0.191

0.451

P12

3.63×10−3

6.68×10−5

3.97×10−3

0.180

0.445

  1. Note that in the case of geneSIMMS, each biomarker size is considered separately