D | H | H_{fw} | K | K_{d} | ||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|

SNM | BN | SNM | BN | SNM | BN | SNM | BN | SNM | BN | |||||||||||

B | S | B | S | B | S | B | S | B | S | B | S | B | S | B | S | B | S | B | S | |

10 | 81 | 78 | 0 | 0 | 28 | 28 | 0 | 0 | 0 | 0 | 0 | 0 | 110 | 106 | 0 | 0 | 27* | 27* | 0 | 0 |

7 | 80 | 78 | 12 | 0 | 27 | 25 | 0 | 1 | 0 | 0 | 0 | 0 | 106 | 101 | 0 | 0 | 24* | 24* | 0 | 0 |

1 | 55 | 28 | 0 | 0 | 13 | 11 | 0 | 0 | 0 | 0 | 0 | 0 | 55 | 55 | 0 | 0 | 14* | 115 | 0 | 0 |

0 | 1 | 1 | 0 | 0 | 7 | 1 | 2 | 2 | 0 | 0 | 0 | 0 | 39 | 30 | 0 | 0 | 57 | 57 | 0 | 0 |

Outlier loci were determined by the calculation of two-sided

*P*-values and then by controlling the false discovery rate (FDR) at 0.05. The FDR method implemented in qvalue (Storey 2002) can use either a bootstrapping or a smoothing method while estimating parameters. The results of both methods are presented here (*B*and*S*, respectively). SNM, standard neutral model; BN, bottleneck. The bottlenecks are simulated from the joint posterior distributions obtained with the values of indicated. In some cases, the distribution of*P*-values is too skewed, resulting in a negative point estimate of the fraction of*P*-values that come from the null model. In this case, a Bonferroni correction is applied instead, reporting the number of loci with a*P*-value of ≤0.05/115. Such values are marked with an asterisk.