TABLE 2

Estimates of QTL allele effects, nonnormal error

Estimate of α
Gen (%)DSMGLRML
Simulated value = 0.1, mixture error
1000.101 ± 0.0030.101 ± 0.0030.101 ± 0.0030.101 ± 0.003
500.106 ± 0.0030.108 ± 0.0030.108 ± 0.0030.108 ± 0.003
100.114 ± 0.0040.128 ± 0.0050.132 ± 0.0050.131 ± 0.005
Simulated value = 0.2, mixture error
1000.201 ± 0.0030.201 ± 0.0030.200 ± 0.0030.201 ± 0.003
500.210 ± 0.0030.214 ± 0.0030.214 ± 0.0030.215 ± 0.003
100.221 ± 0.0040.250 ± 0.0050.263 ± 0.0050.270 ± 0.006
Simulated value = 0.1, χ2-error
1000.100 ± 0.0030.100 ± 0.0030.101 ± 0.0030.100 ± 0.003
500.126 ± 0.0030.120 ± 0.0030.132 ± 0.0030.121 ± 0.003
100.210 ± 0.0050.131 ± 0.0050.228 ± 0.0060.133 ± 0.005
Simulated value = 0.2, χ2-error
1000.200 ± 0.0030.200 ± 0.0030.205 ± 0.0030.200 ± 0.003
500.241 ± 0.0030.228 ± 0.0030.253 ± 0.0030.232 ± 0.003
100.335 ± 0.0030.207 ± 0.0040.422 ± 0.0050.216 ± 0.004
Simulated value = 0.5, χ2-error
100.382 ± 0.0020.327 ± 0.0030.526 ± 0.0060.359 ± 0.004

Estimates are means and standard errors of 100 replicates, where Gen refers to the percentage of animals genotyped. Estimation methods were those of Darvasi and Soller (1992; DS) and Muranty and Goffinet (1997; MG). LR, logistic regression; ML, maximum likelihood. Error terms simulated were a mixture of normal distributions and χ2 with 4 d.f.