Skip to main content

Table 2 Impact of genotyping error (rate = ϵ) on type I error and power

From: Influence of genotyping error in linkage mapping for complex traits – an analytic study

Error rate ϵ Selection ρ EC ED I MathType@MTEF@5@5@+=feaagaart1ev2aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacPC6xNi=xH8viVGI8Gi=hEeeu0xXdbba9frFj0xb9qqpG0dXdb9aspeI8k8fiI+fsY=rqGqVepae9pg0db9vqaiVgFr0xfr=xfr=xc9adbaqaaeGaciGaaiaabeqaaeqabiWaaaGcbaWenfgDOvwBHrxAJfwnHbqeg0uy0HwzTfgDPnwy1aaceaGae8heHKeaaa@3691@
    Power Type I Error × 10-4 Power Type I Error × 10-4 Power Type I Error × 10-4
0.005 1% 0.1 0.87 0.6 0.92 1.6 0.90 1.1
   0.2 0.87 0.6 0.92 1.5 0.91 1.3
   0.3 0.87 0.5 0.91 1.5 0.91 1.3
   0.4 0.87 0.5 0.91 1.4 0.91 1.3
   0.5 0.87 0.5 0.91 1.4 0.91 1.3
   0.6 0.87 0.5 0.91 1.3 0.91 1.2
  10% 0.1 0.85 0.4 0.93 2.2 0.90 1.1
   0.2 0.85 0.4 0.93 2.0 0.91 1.2
   0.3 0.84 0.3 0.92 1.9 0.91 1.3
   0.4 0.84 0.3 0.92 1.7 0.91 1.4
   0.5 0.84 0.3 0.92 1.6 0.91 1.4
   0.6 0.85 0.4 0.91 1.5 0.91 1.3
  30% 0.1 0.83 0.3 0.94 2.8 0.90 1.1
   0.2 0.82 0.2 0.93 2.4 0.90 1.1
   0.3 0.82 0.2 0.93 2.1 0.91 1.2
   0.4 0.82 0.2 0.92 1.9 0.91 1.2
   0.5 0.83 0.3 0.92 1.6 0.91 1.2
   0.6 0.85 0.4 0.91 1.5 0.91 1.2
0.01 1% 0.1 0.84 0.3 0.93 2.4 0.91 1.3
   0.2 0.83 0.3 0.93 2.2 0.92 1.7
   0.3 0.83 0.3 0.93 2.1 0.92 1.8
   0.4 0.83 0.3 0.92 1.9 0.92 1.7
   0.5 0.83 0.3 0.92 1.7 0.92 1.6
   0.6 0.83 0.3 0.92 1.6 0.91 1.5
  10% 0.1 0.78 0.1 0.95 4.5 0.91 1.3
   0.2 0.78 0.1 0.95 3.9 0.91 1.5
   0.3 0.77 0.1 0.94 3.4 0.92 1.7
   0.4 0.77 0.1 0.94 2.9 0.92 1.9
   0.5 0.77 0.1 0.93 2.5 0.92 1.9
   0.6 0.78 0.1 0.93 2.1 0.92 1.8
  30% 0.1 0.73 0.1 0.96 7.1 0.90 1.2
   0.2 0.71 0.1 0.96 5.6 0.91 1.3
   0.3 0.71 0.0 0.95 4.4 0.91 1.4
   0.4 0.71 0.1 0.94 3.4 0.91 1.5
   0.5 0.74 0.1 0.93 2.6 0.91 1.5
   0.6 0.78 0.1 0.93 2.1 0.91 1.5
  1. Impact of genotyping error (rate = ϵ) on power and type I error of linkage test in selective designs (indexed by degree of Selection and sib-sib trait correlation ρ) – Nominal error rate = 10-4, QTL effect γ = 0.1 and sample size equivalent to a Fisher's information = 2500 in each design (provides 90% power in absence of genotyping error)