Advances in Computational Methods for Phylogenetic Networks in the Presence of Hybridization

Abstract

Phylogenetic networks extend phylogenetic trees to allow for modeling reticulate evolutionary processes such as hybridization. They take the shape of a rooted, directed, acyclic graph, and when parameterized with evolutionary parameters, such as divergence times and population sizes, they form a generative process of molecular sequence evolution. Early work on computational methods for phylogenetic network inference focused exclusively on reticulations and sought networks with the fewest number of reticulations to fit the data. As processes such as incomplete lineage sorting (ILS) could be at play concurrently with hybridization, work in the last decade has shifted to computational approaches for phylogenetic network inference in the presence of ILS. In such a short period, significant advances have been made on developing and implementing such computational approaches. In particular, parsimony, likelihood, and Bayesian methods have been devised for estimating phylogenetic networks and associated parameters using estimated gene trees as data. Use of those inference methods has been augmented with statistical tests for specific hypotheses of hybridization, like the D-statistic. Most recently, Bayesian approaches for inferring phylogenetic networks directly from sequence data were developed and implemented. In this chapter, we survey such advances and discuss model assumptions as well as methods' strengths and limitations. We also discuss parallel efforts in the population genetics community aimed at inferring similar structures. Finally, we highlight major directions for future research in this area.

Species Tree Inference on Data with Paralogs is Accurate Using Methods Intended to Deal with Incomplete Lineage Sorting

“…Zhu et al, 2016), but, to the best of our knowledge, such studies under scenarios that incorporate all the aforementioned processes do not exist yet. It is important to highlight, as well, that great strides have been made in developing methods for phylogenetic network inference in the presence of ILS (Elworth et al, 2018), but none currently incorporate gene duplication and loss. We believe methods along the lines described in the previous section could be promising for accurate and scalable phylogenomic inferences without sacrificing much of the data.…”

A Divide-and-Conquer Method for Scalable Phylogenetic Network Inference from Multi-locus Data

“…Given that computational complexity of Bayesian inference of trinets , we focus our attention here on a subset of 24 phylogenetic networks that we sampled to reflect varying complexity levels. As discussed in (Zhu et al, 2016;Elworth et al, 2018), the complexity of phylogenetic networks arises not only from the number of leaves or number of reticulation nodes, but also in how the reticulation nodes are structured in the network.To allow for a careful assessment of the accuracy of our approach, we define a simple complexity measure of networks as follows. We define the complexity of Ψ as r∈R(Ψ) |L(r)| + |L(p 1 (r))| + |L(p 2 (r))| + |X | · |AR Ψ (r)|, where L(u) is the set of leaves under node u, and p 1 (u) and p 2 (u) are the two parents of reticulation node u.…”

“…However, these methods suffer from several major performance bottlenecks. Methods that evaluate the full likelihood (all of the aforementioned methods, except for the pseudo-likelihood method of Yu and Nakhleh (2015)) suffer from the prohibitive computational requirements of likelihood calculations Elworth et al, 2018). Currently, computing network likelihood is feasible only for fewer than 10 species and a very small number of reticulations.…”

The Timing and Direction of Introgression Under the Multispecies Network Coalescent

“…To address the aforementioned problems, here we develop an explicit model for the timing and direction of introgression based on the multispecies network coalescent (Yu et al 2012(Yu et al , 2014Wen et al 2016b). The multispecies network coalescent model generalizes the multispecies coalescent (Hudson 1983;Rannala and Yang 2003) to allow for both incomplete lineage sorting and introgression (reviewed in Degnan 2018;Elworth et al 2018). Under this model, a single sample taken from each of the extant lineages traces its history back through the network, following alternative paths produced by reticulations with a probability proportional to the amount of introgression that has occurred.…”