TABLE 6

The effect of missing data on the power of detection of QTL × E interaction and accuracy of QTL location when the number of environments is large

SiModelNlinNenvEmbedded Image L(cM)α%→5βet (%) 10.1
S1MA32001020066.0 ± 2.03674116
2001005059.7 ± 0.94948260
20010010060.0 ± 0.451009998
S3MA3100504060.0 ± 0.721009893
200504060.2 ± 0.56989688
S4MA1200504061.4 ± 1.10816946
MA2200504061.0 ± 1.10897756
MA3200504060.0 ± 0.72989382
S4MA12005010060.1 ± 0.41989793
MA22005010060.2 ± 0.39999897
MA32005010059.9 ± 0.27100100100
  • Nlin is the total number of genotypes (lines) in the mapping population, Nenv is the number of environments, and Embedded Image is the mean number of genotypes scored per environments. To generate the data, the initial form of the dependence of the QTL on environment was used as presented in Table 1, but with a correspondingly smaller step in changes of the independent variable.