The Cheshire cat reloaded: causes and consequences of the evolution of dormancy/quiescence in parasites.

在许多植物、动物和微生物(包括寄生虫或病原体)中，休眠和休眠作为一种更好的进化策略在不可预测的环境中进化。该项目的中心假设是，宿主和寄生虫之间的共同进化促进了这种条件的产生。这种进化情景被称为柴郡猫假说(对宿主或寄生虫而言)。我们之前已经证明，1)种子库和休眠可以由于共同进化动力学而在宿主中进化，以及2)种子库本身可以减缓甚至抑制共同进化的振荡，这取决于种子萌发的功能。由于等位基因频率振荡对促进更好的对冲进化至关重要，休眠进化和共同进化动态之间发生了复杂的生态进化相互作用。该项目新阶段的目标是双重的。首先，我们想要了解和预测寄生虫进化休眠或静止作为共同进化更好的对冲策略的条件。其次，我们希望开发模拟和统计推理工具，允许使用寄生虫全基因组数据来估计休眠/静止参数。在第一个目标中，我们研究了投注套期保值在共同进化动力学下演化的条件。建立了两种休眠模型：短休眠和超长休眠。共同进化是在两个宿主和两种寄生虫类型的连续时间流行病学模型中产生的，带有寄生虫休眠/休眠。我们利用自适应动力学的分析方法，推导出了两种休眠模型在不同条件下休眠/休眠的ESS的解析解，并对两种休眠模型的时间尺度进行了分离。在第二个目标中，我们希望开发工具来利用寄生虫的全基因组数据来检测和估计休眠/静止的存在。因此，随着时间的推移，将为有性或无性繁殖物种开发弱休眠/强休眠/静止的全基因组模拟器。这些模拟器将被集成到具有深度学习(卷积神经网络)的近似贝叶斯计算(ABC)框架中，以基于全基因组数据执行休眠/静止参数的统计推断。在第三个目标中，我们将应用我们的ABC方法来研究几个生物学上怀疑静止/休眠但未经测试的寄生虫物种的全基因组：人类寄生虫(间日疟原虫、恶性疟原虫、芽孢杆菌、结核分枝杆菌)、大型水蚤的寄生虫和车前草的白粉病。因此，我们的总体目标是首次了解由于共同进化而导致的寄生虫物种休眠或静止进化的原因和后果，并发展一种新的种群基因组推断方法来揭示有性寄生虫和无性寄生虫物种休眠的存在和估计休眠率。

Quiescence and dormancy evolve as a bet-hedging evolutionary strategies in many plants, animals, and micro-organisms (including parasites or pathogens) in unpredictable environments. The central hypothesis of the project states that coevolution between hosts and their parasites promotes such conditions. This evolutionary scenario is referred to as the Cheshire Cat hypothesis (for either hosts or parasites). We have previously shown that 1) seed banking and dormancy can evolve in hosts due to coevolutionary dynamics, and 2) depending on the germination function, the seed bank itself can slow down or even damp off the coevolutionary oscillations. As allele frequency oscillations are critical to promote bet-hedging evolution, a complex eco-evolutionary interplay occurs between the evolution of dormancy and coevolutionary dynamics. The goals of this new phase of the project are twofold. First we want to understand and predict the conditions for parasites to evolve dormancy or quiescence as bet-hedging strategies in response to coevolution. Second, we want to develop simulation and statistical inference tools which allow to estimate the parameter of dormancy/quiescence using parasite full genome data. In the first objective, we study the conditions for bet-hedging to evolve under coevolutionary dynamics. Two dormancy models are developed: a short and a very long dormancy. Coevolution is generated within a continuous-time epidemiological model with two host and two parasite types with parasite dormancy/quiescence. We use analytical methods of adaptive dynamics to derive analytical solutions for the ESS of dormancy/quiescence under different conditions and separation of time scales for both dormancy models. In the second objective, we want to develop tools for detecting and estimating the existence of dormancy/quiescence using parasite full genome data. Therefore, forward in time full genome simulators of weak and strong dormancy/quiescence will be developed for sexually or asexually reproducing species. These simulators will be integrated into an Approximate Bayesian Computation (ABC) framework with Deep Learning (Convolutional Neural Networks) to perform statistical inference of the dormancy/quiescence parameters based on full genome data. In the third objective, we will apply our ABC methods to full genomes of several parasite species for which quiescence/dormancy is biologically suspected but untested: human parasites (Plasmodium vivax, P. falciparum, Bacillus sp., Mycobacter tuberculosis), the parasites of Daphnia magna, and the powdery mildewof Plantago lanceolata. Our general aim is thus to provide the first understanding of the causes and consequences of the evolution of dormancy or quiescence in parasite species due to coevolution, and to develop a novel population genomic inference method to reveal the existence and estimate the rate of dormancy in sexual and asexual parasite species.