Skip to main content

Table 2 Estimates of Type I error for MBMDR-PC, MBMDR-GC, and MDR-SP, with a nominal 0.05 FWER level

From: Performance of model-based multifactor dimensionality reduction methods for epistasis detection by controlling population structure

 

b = 40%

Method

Sample sizes

Markers

d = 0.1

d = 0.3

MBMDR-PC

200

200

0.050

0.046

  

400

0.051

0.051

  

800

0.046

0.048

 

300

200

0.046

0.047

  

400

0.047

0.049

  

800

0.049

0.047

 

400

200

0.050

0.053

  

400

0.051

0.054

  

800

0.048

0.046

MDR-SP

200

200

0.054

0.054

  

400

0.062

0.055

  

800

0.062

0.050

 

300

200

0.051

0.056

  

400

0.055

0.051

  

800

0.044

0.046

 

400

200

0.044

0.050

  

400

0.046

0.052

  

800

0.044

0.065

MBMDR-GC

200

200

0.059

0.058

  

400

0.061

0.065

  

800

0.064

0.067

 

300

200

0.060

0.065

  

400

0.062

0.071

  

800

0.063

0.068

 

400

200

0.066

0.069

  

400

0.061

0.071

  

800

0.063

0.066

  1. Note: d denotes the difference of candidate allele frequencies in the two subpopulations, and b denotes the percentage of cases from the European subpopulation