Organismal phenotypes usually have a quantitative distribution, and their genetic architecture can be studied by genome-wide association (GWA) mapping approaches. In most of such studies, it has become clear that many genes of moderate or small effects contribute to the phenotype.1, 2, 3, 4 Hence, the attention has turned toward the loci falling below the GWA cut-off, which may contribute to the phenotype through modifier interactions with a set of core genes, as proposed in the omnigenic model.⁵ One can thus predict that both moderate effect GWA-derived candidate genes and randomly chosen genes should have a similar likelihood to affect a given phenotype when they are analyzed via gene disruption assays. We have tested this hypothesis by using an automated phenotyping system for Drosophila pupal phenotypes.^6,7 We first identified candidate genes for pupal length in a GWA based on the Drosophila Genetic Reference Panel (DGRP)^8,9 and showed that most of these candidate genes are indeed involved in the phenotype. We then randomly chose genes below a GWA significance threshold and found that three-quarters of them had also an effect on the trait with comparable effect sizes as the GWA candidate genes. We further tested the effects of these knockout lines on an independent behavioral pupal trait (pupation site choice) and found that a similar fraction had a significant effect as well. Our data thus confirm the implication that a large number of genes can influence independent quantitative traits.

有机体表型通常具有定量分布，它们的遗传结构可以通过全基因组关联 (GWA) 作图方法进行研究。在大多数此类研究中，很明显，许多中等或小影响的基因对表型有贡献。1、2、3、4 因此，注意力已转向低于 GWA 截止值的基因座，这可能有助于如全基因模型中所提出的，通过修饰剂与一组核心基因的相互作用来改变表型。⁵因此，人们可以预测，当通过基因破坏分析对其进行分析时，中等效应 GWA 衍生的候选基因和随机选择的基因应该具有相似的影响给定表型的可能性。我们通过使用果蝇的自动表型系统测试了这一假设蛹表型。^{6 , 7}我们首先根据果蝇遗传参考小组 (DGRP) ^{8 、9}在 GWA 中确定了蛹长度的候选基因，并表明这些候选基因中的大多数确实与表型有关。然后，我们随机选择低于 GWA 显着性阈值的基因，发现其中四分之三的基因对性状也有影响，其影响大小与 GWA 候选基因相当。我们进一步测试了这些敲除系对独立的蛹行为特征（化蛹位点选择）的影响，发现类似的部分也有显着影响。因此，我们的数据证实了大量基因可以影响独立数量性状的暗示。