Skip to main content

Table 3 Model-wised performances comparison at pan-cancer level (12 TCGA (The Cancer Genome Atlas) cancer types) by pairwise paired t-test (A) and linear mixed-effects models test (B), according to metrics concordance index and p-value of log-rank test. Note that for concordance index, larger t-statistic/coefficient indicated better performance at pan-cancer level, while the p-value of log-rank test was on the contrary

From: Deep learning-based cancer survival prognosis from RNA-seq data: approaches and evaluations

(A) Pairwise Paired T-test

 

Distribution 2

DeepSurv

AECOX

t

P

t

P

Distribution 1

Cox-nnet

concordance index

3.1843

2.32E-03

3.2281

2.04E-03

p-value of log-rank test

−1.4006

1.67E-01

−0.8962

3.74E-01

DeepSurv

concordance index

–

–

−0.6732

5.03E-01

p-value of log-rank test

–

–

0.5164

6.07E-01

Notes: t denotes the pairwise paired Student’s t-test statistic, P denotes the p-value obtained.

(B) Linear Mixed-Effects Models Test

 

Distribution 2

DeepSurv

AECOX

β

P

β

P

Distribution 1

Cox-nnet

concordance index

0.0195

1.97E-02

0.0142

1.12E-01

p-value of log-rank test

−0.0489

2.52E-01

−0.0294

4.85E-01

DeepSurv

concordance index

–

–

−0.0052

5.85E-01

p-value of log-rank test

–

–

0.0195

6.62E-01

  1. Notes: β denotes the coefficient (slope) of linear mixed-effects models, P denotes the p-value obtained.