Table 1 Concordance between the LD clumped regions discovered by BayesW or fastGWA.

From: Genomic architecture and prediction of censored time-to-event phenotypes with a Bayesian genome-wide analysis

 

BayesW: no

BayesW: no

BayesW: yes

BayesW: yes

Total BayesW

Phenotype

fastGWA: no

fastGWA: yes

fastGWA: no

fastGWA: yes

 

Time to Angina

290,220

0

128

0

128

Time to Heart attack

289,787

0

128

0

128

Time to HBP

291,674

4

653

10

666

Time to Menarche

292,127

242

223

191

414

Time to Menopause

292,202

125

126

97

227

Time to Diabetes

290,599

40

174

8

183

  1. We split the genome into LD clumped regions and we evaluated the significance of each of the regions using the results from the groups BayesW model and the fastGWA model. The fastGWA results for our CAD and T2D definition were missing so instead time-to-angina and time-to-heart attack are shown for CAD and time-to-diabetes is shown for T2D. Here, BayesW calls an LD clumped region significant if the PPWV of the region (explaining at least 0.001% of the genetic variance) is higher than 0.9; fastGWA calls an LD clumped region significant if there exists at least one marker with a p-value < 5 × 10−8. We find that although for age-at-menarche and age-at-menopause there exists an abundance of regions with concordant significance, for other traits most of the discovered regions differ between two methods. For creating the comparison only overlapping markers were used; in the column Total BayesW we show the total number of discovered LD clumped regions, including those that did not have a counterpart among fastGWA results.
Back to article page