Table 3 SNP-based heritabilities for GWAS and WGWAS

From: Correcting for volunteer bias in GWAS increases SNP effect sizes and heritability estimates

Phenotype

GWAS

WGWAS

P

GWAS

WGWAS

 

h2 (SE)

h2 (SE)

 

Intercept (SE)

Intercept (SE)

Age at First Birth

0.1657 (0.0073)

0.2135 (0.0143)

1.28 × 10−5

1.035 (0.010)

1.015 (0.008)

BMI

0.2281 (0.0065)

0.2381 (0.0091)

0.14

1.127 (0.015)

1.033 (0.011)

Breast cancer

0.0259 (0.0034)

0.0512 (0.0059)

2.37 × 10−8

1.021 (0.008)

0.985 (0.007)

Drinks per Week

0.0599 (0.0030)

0.0739 (0.0054)

7.44 × 10−4

1.005 (0.008)

0.985 (0.006)

Height

0.4235 (0.0189)

0.4464 (0.0206)

0.059

1.479 (0.035)

1.169 (0.020)

Physical activity

0.0281 (0.0019)

0.0311 (0.0044)

0.408

0.996 (0.007)

0.993 (0.007)

Self-rated health

0.0972 (0.0029)

0.1250 (0.0052)

9.35 × 10−13

1.052 (0.010)

1.009 (0.008)

Severe Obesity

0.0416 (0.0022)

0.0584 (0.0045)

1.83 × 10−6

1.017 (0.008)

0.995 (0.008)

Type 1 Diabetes

0.0054 (0.0014)

0.0432 (0.0035)

1.63 × 10−41

1.019 (0.007)

0.940 (0.006)

Years of Education

0.1482 (0.0052)

0.1775 (0.0073)

2.07 × 10−9

1.164 (0.016)

1.053 (0.011)

  1. SNP-based heritabilities for GWAS (column 1) and WGWAS (column 2) were estimated using LD-score regression (see Methods). The third column shows the p-value for the null hypothesis that the GWAS and WGWAS heritabilities are the same, based on a two-sided Z-test (see Methods). The fourth and fifth columns show the intercept of the LD-score regression in GWAS and WGWAS, respectively. An intercept  > 1 can be attributed to bias arising from population stratification22.
Back to article page