Evolutionary genetic studies have uncovered abundant evidence for genomic hotspots of phenotypic evolution, as well as biased patterns of mutations at those loci. However, the theoretical basis for this concentration of particular types of mutations at particular loci remains largely unexplored. In addition, historical contingency is known to play a major role in evolutionary trajectories, but has not been reconciled with the existence of such hotspots. For example, do the appearance of hotspots and the fixation of different types of mutations at those loci depend on the starting state and/or on the nature and direction of selection? Here, we use a computational approach to examine these questions, focusing the anthocyanin pigmentation pathway, which has been extensively studied in the context of flower color transitions. We investigate two transitions that are common in nature, the transition from blue to purple pigmentation and from purple to red pigmentation. Both sets of simulated transitions occur with a small number of mutations at just four loci and show strikingly similar peaked shapes of evolutionary trajectories, with the mutations of the largest effect occurring early but not first. Nevertheless, the types of mutations (biochemical vs. regulatory) as well as their direction and magnitude are contingent on the particular transition. These simulated color transitions largely mirror findings from natural flower color transitions, which are known to occur via repeated changes at a few hotspot loci. Still, some types of mutations observed in our simulated color evolution are rarely observed in nature, suggesting that pleiotropic effects further limit the trajectories between color phenotypes. Overall, our results indicate that the branching structure of the pathway leads to a predictable concentration of evolutionary change at the hotspot loci, but the types of mutations at these loci and their order is contingent on the evolutionary context.

中文翻译：

---

结构和偶然性决定花色演化的突变热点


进化遗传学研究发现了表型进化的基因组热点的大量证据，以及这些位点的突变模式。然而，特定基因座上特定类型突变集中的理论基础在很大程度上仍未得到探索。此外，众所周知，历史偶然性在进化轨迹中发挥着重要作用，但尚未与此类热点的存在相一致。例如，热点的出现以及这些基因座上不同类型突变的固定是否取决于起始状态和/或选择的性质和方向？在这里，我们使用计算方法来研究这些问题，重点关注花青素色素沉着途径，该途径已在花朵颜色转变的背景下进行了广泛研究。我们研究了自然界中常见的两种转变，即从蓝色到紫色色素沉着的转变以及从紫色到红色色素沉着的转变。两组模拟转变仅在四个基因座上发生少量突变，并显示出惊人相似的进化轨迹峰值形状，其中最大效应的突变发生较早，但不是首先出现。然而，突变的类型（生化突变与监管突变）及其方向和幅度取决于特定的转变。这些模拟的颜色转变在很大程度上反映了自然花朵颜色转变的发现，众所周知，自然花朵颜色转变是通过一些热点基因座的重复变化而发生的。尽管如此，在我们模拟的颜色进化中观察到的某些类型的突变在自然界中很少观察到，这表明多效性效应进一步限制了颜色表型之间的轨迹。 总的来说，我们的结果表明，该途径的分支结构导致热点基因座处的进化变化可预测集中，但这些基因座的突变类型及其顺序取决于进化背景。