TABLE 4

Simulation of chromosome-wise type I error α for IM in a BC1 population

Empirical type I errorsThresholds by
Rebaïet al. (1994)
Step
size
Quick methodRebaïet al. (1994)
MaDistributionaα = 0.01α = 0.05α = 0.01α = 0.05α = 0.01α = 0.05
100N(0, 1)10.00560.03060.00530.027812.979.76
50N(0, 1)10.00800.03310.00770.032512.309.09
20N(0, 1)10.00820.04150.00880.043711.428.22
10N(0, 1)10.00850.04190.00990.046410.767.58
5N(0, 1)10.00750.04120.00960.049210.106.93
5N(0, 1)20.00650.04180.00850.049110.106.93
5N(0, 1)50.00740.04020.00940.046010.106.93
100U(0, 1)10.00730.03180.00680.029612.979.76
100χ2110.00360.02620.00300.020412.979.76
20χ2110.00470.03160.00540.032011.428.22
  • Equal spacing of markers is assumed. Simulation is based on 10,000 replications. Length of chromosome = 100 cM.

  • a M, No. of intervals on a chromosome.

  • b N(0, 1), standard normal; U(0, 1), uniform; Formula , chi square with 1 d.f.