Table 1 Damaging PALB2 missense variants associate with increased breast cancer risk

From: Site-saturation functional screens identify PALB2 missense variants associated with increased breast cancer risk

BRIDGES CARRIERS CZECANCAa

Category

Cases/controls

Odds ratio

CI (95%)b

P-value

 

Functional

463/379

1.07

0.93–1.23

0.333

 

Intermediate

159/109

1.28

1.00–1.64

0.050

 

Damaging

16/3

4.67

1.36–18.82

0.009

 

Standardsc

220/38

5.18

3.68–7.38

4.17 × 10−28

UK Biobank

Category

Cases/controls

Odds ratio

CI (95%)

P-value

 

Functional

59/587

0.99

0.75–1.30

1.000

 

Intermediate

34/206

1.63

1.12–2.34

0.013

 

Damaging

3/6

4.94

1.07–20.68

0.043

 

Standards

71/170

4.14

3.13–5.50

3.7 × 10−19

Clinical cohort/gnomAD

Category

Cases/controls

Odds ratio

CI (95%)

P-value

 

Functional

245/627

1.10

0.94–1.27

0.228

 

Intermediate

519/264

1.00

0.86–1.16

1.000

 

Damaging

51/8

3.25

1.58–7.51

7.83 × 10−4

 

Standards

919/65

7.22

5.63–9.38

4.76 × 10−93

Clinical cohort/All of Us

Category

Cases/controls

Odds ratio

CI (95%)

P-value

 

Functional

245/396

1.09

0.93–1.28

0.285

 

Intermediate

519/192

0.87

0.73–1.03

0.096

 

Damaging

51/6

2.73

1.16–6.53

0.013

 

Standards

919/54

5.47

4.16–7.24

2.68 × 10−55

  1. aBurden-type case-control association analyses were conducted to evaluate the relationship between functional, intermediate, and damaging missense variants and breast cancer across three cohorts. First, in the BRIDGES, CARRIERS, and CZECANCA population-based studies (combined total of 110,223 cases and 96,204 controls), associations were tested using a two-sided Fisher’s exact tests. Second, in the UK Biobank cohort (20,856 cases and 206,098 controls) associations were tested using Firth logistic regression, adjusting for age and ancestry and deriving P-values from the likelihood ratio of the Firth logistic regression. Third, a clinical cohort of 356,861 breast cancer cases tested for PALB2 variants by Ambry Genetics was compared to two control groups, gnomAD v4.1 non-UK Biobank female controls (n = 181,922) and All of Us Research Program non-cancer female controls (n = 114,489), using Firth logistic regression, adjusting for age and ancestry and deriving P-values from the likelihood ratio of the Firth logistic regression.
  2. bCI confidence interval.
  3. cStandards: nonsense variants of PALB2 observed in each case control study.