Skip to main content

Table 2 Significance level (α) and type I error rate for Q1 and Q2

From: Improved power by collapsing rare and common variants based on a data-adaptive forward selection strategy

Trait

T ind

T sum

T ws

FSind

FSsum

FSws

T com

Q1: α

0.042

0.038

0.0171

0.052

0.035

0.003

0.038

Q1: type I

6%

6%

6%

5.5%

6%

6%

6%

Q2: α

0.041

0.03

0.103

0.06

0.054

0.083

0.053

Q2: type I

5%

5%

5%

5%

5%

5%

5%

  1. Tind, Tsum, and Tws use adjusted p-values by the Bonferroni correction; FSind, FSsum, FSws, and Tcom use the global permutation p-value. The power of tests is calculated on the basis of the same significance levels as the type I error.