Skip to main content

Table 2 Empirical levels for different nominal alpha levels (0.01 and 0.05) using different reference data sets

From: Quick, “Imputation-free” meta-analysis with proxy-SNPs

Strategya

Referenceb

Modelc

α = 0.05d

α= 0.01e

MA-pointwise

-

fixed

0.047

0.009

 

-

random

0.037

0.006

MA-proxy

ideal

fixed

0.046

0.009

  

random

0.036

0.007

 

incomplete

fixed

0.046

0.009

  

random

0.036

0.006

 

mismatched

fixed

0.047

0.009

  

random

0.036

0.007

MA-impute

ideal

fixed

0.042

0.007

  

random

0.035

0.006

 

incomplete

fixed

0.041

0.007

  

random

0.034

0.006

 

mismatched

fixed

0.039

0.008

  

random

0.032

0.006

  1. aEnumeration using either the naive approach, proxy algorithm or imputation strategy.
  2. bReference data set used for proxy algorithm and imputation.
  3. cCalculation of the allelic effects by the fixed effect or random effect model.
  4. dEmpirical levels for nominal α of 0.05 (average over all SNPs from chr 22 for 100 permutation replicates).
  5. eEmpirical levels for nominal α of 0.01.