Effect^{c} | Inferred mode^{d} | |||||||||
---|---|---|---|---|---|---|---|---|---|---|

Simulated QTL | Power^{a} | Position^{b} | 2a | a_{pat} | a_{mat} | d | E_{1} | E_{2} | E_{3} | E_{4} |

QTL1 (LG1: 25 cM, E_{4}; a_{mat} = −0.25) | 0.85 | 25.4 (4.1) | — | — | −0.25 (0.04) | — | 0.00 (0.02) | 0.02 (0.19) | 0.00 (0.01) | 0.83 (0.79) |

QTL2 (LG1: 75 cM, E_{1}; 2a = 0.5, d = 0.0) | 0.98 | 75.4 (3.5) | 0.48 (0.07) | — | — | 0.03 (0.11) | 0.85 (0.60) | 0.11 (0.35) | 0.02 (0.03) | 0.00 (0.02) |

QTL3 (LG2: 65 cM, E_{3}; a_{pat} = 0.2) | 0.79 | 61.7 (9.0) | — | 0.20 (0.04) | — | — | 0.02 (0.05) | 0.04 (0.16) | 0.73 (0.77) | 0.00 (0.03) |

Simulated data sets in case II were analyzed by a Bayesian method with the prior and transition probabilities for QTL expression modes modified as uniform, where prior probabilities for

*E*_{1},*E*_{2},*E*_{3}, and*E*_{4}were 0.25 and transition probabilities among expression modes were set at δ = 0.5, κ = 0.25, ω = 0.25, η = 0.5, and ζ = 0.5. The proportions of MCMC iterations fitting QTL1 and QTL2 simultaneously in each analysis were averaged as 0.87 ± 0.22.↵

*a*Proportion of replicates that successfully detected each QTL with a genomewide significance level of*P*= 0.05 in a total of 100 replicates.↵

*b*Average of estimated QTL positions calculated in the replicates with successful detection of QTL with standard deviations given in parentheses.↵

*c*Average of estimated QTL effects corresponding to the simulated expression mode over the replicates with successful QTL detection with standard deviations given in parentheses.↵

*d*Proportion of replicates supporting each QTL expression mode as the most probable state in the replicates with successful detection of QTL. Averages of posterior probabilities of QTL expression modes are given in parentheses for the Bayesian method.