Skip to main content
. 2010 Feb 1;11(4):385–393. doi: 10.1093/bib/bbp070

Table 1:

Simulation study: median computed based on 500 replicates

True model Correlation Assumed model Positive True positive P-value
Cox 0.1 Cox 27 15 < 0.0001
Additive 36 4 0.153
AFT 50 5 0.306
0.6 Cox 18 15 < 0.0001
Additive 38 8 < 0.001
AFT 40 4 0.020
Additive 0.1 Cox 40 5 0.418
Additive 25 13 0.023
AFT 10 0 0.731
0.6 Cox 40 5 0.231
Additive 22 13 0.012
AFT 10 0 0.402
AFT 0.1 Cox 27 10 < 0.0001
Additive 25 8 0.006
AFT 26 14 < 0.0001
0.6 Cox 17 9 < 0.001
Additive 13 5 0.002
AFT 22 15 < 0.0001

Positive: number of genes identified; true positive: number of identified genes truly associated with disease; P-value: P-value of the logrank statistic.