Molecular evidence supports a genic capture resolution of the lek paradox

Allele frequency data for populations of Drosophila melanogaster (columns 6:27). Populations were artificially selected for success (S) or failure (F) in male mating success for 14 generations. There were also three control (C) populations maintained with random males. Data are presented as the counts of the rare (q) and common (p) allele for each population.<br>DiffStat scores were calculated as the minimum allele frequency difference between any combination of success- versus failure-selected line, and is zero unless all four lines of one selection regimen have higher or lower allele frequencies than all four lines of the other selection regimen.<br><br>DiffStat loci are considered significant if the DiffStat (i.e. the minimum difference between regimens) is larger than variation caused by drift (calculated as the maximum allele frequency difference within any regimen).<br>q values were obtained by running quasibinomial GLMs for each locus (success- versus failure-selected), and correcting for multiple testing.<br>The type of polymorphism and gene names for each locus are also presented.

黑腹果蝇（Drosophila melanogaster）种群的等位基因频率数据对应第6至27列。实验种群经14代人工选育，以雄性交配成功率为筛选指标，分为成功组（S组）与失败组（F组）；另设3个对照组（C组）种群，采用随机雄性交配的方式维持种群。数据以各种群的稀有等位基因（q）与常见等位基因（p）的计数形式呈现。
DiffStat得分的计算方法为：成功选育组与失败选育组的任意品系组合间的最小等位基因频率差；仅当某一选育方案的全部4个品系的等位基因频率均高于或低于另一选育方案的全部4个品系时，该得分不为零，否则得分为0。

当DiffStat得分（即两组选育方案间的最小频率差）大于遗传漂变导致的变异（该变异以任一选育方案内的最大等位基因频率差计算）时，该位点被视为具有统计学显著性。q值通过对每个位点（成功选育组与失败选育组）拟合准二项广义线性模型（quasibinomial GLMs）并进行多重检验校正后获得。各位点的多态性类型与基因名称也一并提供。