Skip to main content

Table 1 Type I errors and power on effect of the proportion of causal SNVs ( d% ) under significance level of 0.05

From: The power comparison of the haplotype-based collapsing tests and the variant-based collapsing tests for detecting rare variants in pedigrees

   

# of variants

maxH

hPDT

hPDT-t

maxV

vPDT

vPDT-t

Type I error

d%=0

r%=0

21.82(sd=4.63)

0.053

0.050

0.035

0.045

0.052

0.041

Power

r%=1

d%=0.5

23.02(sd=4.90)

0.456

0.906

0.275

0.465

0.894

0.326

  

d%=0.4

22.63(sd=4.64)

0.373

0.734

0.230

0.396

0.744

0.242

  

d%=0.3

22.65(sd=5.11)

0.316

0.515

0.163

0.306

0.525

0.177

  

d%=0.2

22.31(sd=4.56)

0.261

0.265

0.096

0.251

0.272

0.132

  

d%=0.1

21.89(sd=4.65)

0.148

0.090

0.073

0.154

0.083

0.065

  1. Note: (1) “# of variants” indicates the mean and standard deviation (sd) values of the number of SNVs in a 20Kb test region on 1000 replicates (10 permutations for each replicate); (2) d% is the proportion of rare variants in the causal region to be causal; (3) r% is the proportion of causal variants increase risk.
  2. Abbreviations: maxH: maximal haplotype test; hPDT: haplotype-based pedigree disequilibrium test; hPDT-t: haplotype-based pedigree disequilibrium test with training data; maxV: maximal rare variant test; vPDT: variant-based pedigree disequilibrium test; vPDT-t: rare variant based pedigree disequilibrium test with training data.