Skip to main content

Table 1 Significant associations between tumor somatic mutation counts and germline PRS

From: Somatic mutational profiles and germline polygenic risk scores in human cancer

Cancer type

Somatic mutation count

Germline PRSa

Direction of associationb

p valuec

PRAD

SBS1

Age at menarche

−

2.49 × 10−9

PRAD

SBS1

IBD

+

9.04 × 10−6

PRAD

SBS1

CD

+

4.63 × 10−8

PRAD

SBS1

UC

+

3.71 × 10−9

PRAD

SBS1

GBM

−

3.29 × 10−8

PRAD

SBS1

HNSC

−

2.07 × 10−9

PRAD

SBS1

BMI

+

6.09 × 10−8

PRAD

SBS1

Drinks per week

−

1.69 × 10−5

BRCA

APOBEC-relatedd

IBD

+

1.79 × 10−6

BRCA

SBS1

Age at menarche

+

1.47 × 10−5

BRCA ER+

SBS1

Age at menarche

+

2.89 × 10−5

BRCA ER-

SBS1

HNSC

−

1.99 × 10−6

BRCA ER-

TSMC

HNSC

−

5.31 × 10−7

COAD

SBS1

Cigarettes per day

−

1.71 × 10−5

COAD

TSMC

HNSC

−

1.71 × 10−5

GBM

SBS1

OV

+

1.97 × 10−5

UCEC

SBS40

CD

−

7.71 × 10−7

  1. aPRS is calculated from the germline genetic data of the same TCGA patient as the tumor sample
  2. bDirection of the association between somatic mutation count and PRS from zero-inflated negative binomial model, negative binomial model, or linear model adjusting for age at cancer diagnosis, sex, and the top 10 genetic PCs
  3. cP value associated with PRS. P values are obtained from likelihood ratio test of model with PRS and model without PRS. For zero-inflated negative binomial model, the results are from testing the count and logistic model jointly. Age at cancer diagnosis, sex, and the top 10 genetic PCs were adjusted as covariates in all models
  4. dAPOBEC-related signature count is the sum of SBS2 and SBS13 mutation counts, both signatures are attributed to the enzymatic activity of the APOBEC family of cytidine deaminases