Skip to main content

Table 3 The 10 most significant multivariate Cox PH models as measured by log-rank p-value

From: Tumor immunogenomic signatures improve a prognostic model of melanoma survival

Model (adjusted by age and stage at primary diagnosis)

Discovery (N = 139)

P

Validation 2 (N = 69)

P

Pooled (Discovery + Validation 2, N = 208)

P

Meta (N = 278)

P

TMB + MacReg + IFNγRes

1.70E−07

1.28E−02

3.37E−10

8.80E−14

TMB + IFNγRes

1.71E−07

1.74E−02

8.26E−10

2.09E−13

MacReg * IFNγRes + TMB

2.16E−07

1.03E−02

1.12E−09

3.30E−13

TMB + Lymphocyte infiltration signature score + IFNγRes

2.84E−07

1.44E−02

6.87E−10

2.88E−13

TMB + MacReg

2.90E−07

6.66E−03

5.23E−10

2.64E−13

Lymphocyte infiltration signature score * IFNγRes + TMB

3.24E−07

2.51E−02

1.97E−09

1.10E−12

TMB * MacReg + IFNγRes

3.40E−07

2.26E−03

6.04E−10

9.59E−14

TMB * IFNγRes + MacReg

3.53E−07

2.22E−02

1.18E−09

3.46E−13

Lymphocyte infiltration signature score * IFNγRes + TMB + MacReg

4.07E−07

3.80E−02

3.05E−09

1.02E−12

TMB * IFNγRes

4.84E−07

3.16E−02

3.04E−09

9.12E−13

  1. Models were adjusted by age and stage at primary diagnosis with covariates that were associated with survival in univariate Cox PH regressions. Interaction effects between variables were assessed (*). The model that was most significantly associated with survival included MacReg, IFNγRes, and TMB. The full table of results are provided in Additional file 1: Table S5