Data from: A two-state model of tree evolution and its applications to alu retrotransposition

Models of tree evolution have mostly focused on capturing the cladogenesis processes behind speciation. Processes that derive the evolution of genomic elements, such as repeats, are not necessarily captured by these existing models. In this paper, we design a model of tree evolution that we call the dual-birth model, and we show how it can be useful in studying the evolution of short Alu repeats found in the human genome in abundance. The dual-birth model extends the traditional birth-only model to have two rates of propagation, one for active nodes that propagate often, and another for inactive nodes, that with a lower rate, activate and start propagating. Adjusting the ratio of the rates controls the expected tree balance. We present several theoretical results under the dual-birth model, introduce parameter estimation techniques, and study the properties of the model in simulations. We then use the dual-birth model to estimate the number of active Alu elements and their rates of propagation and activation in the human genome based on a large phylogenetic tree that we build from close to one million Alu sequences.

现有树状演化模型大多聚焦于刻画物种形成背后的分支演化（cladogenesis）过程，而诸如重复序列这类基因组元件的演化驱动过程，却未必能被此类现有模型所涵盖。本文中，我们提出了一种名为双生模型（dual-birth model）的树状演化模型，并展示了其在解析人类基因组中大量富集的短Alu重复序列演化规律方面的应用价值。双生模型对传统仅增殖模型进行了拓展，引入两类传播速率：一类对应频繁发生增殖的活跃节点，另一类对应以较低速率完成激活并启动增殖的非活跃节点。通过调整两类速率的比值，可调控模型预期生成的树状结构的平衡性。我们推导得到了双生模型框架下的若干理论结果，提出了参数估计方法，并通过仿真实验分析了该模型的特性。随后，我们基于近百万条Alu序列构建了大型系统发育树（phylogenetic tree），并利用双生模型估算了人类基因组中活跃Alu元件的数量，以及它们的增殖与激活速率。