Inferring explicit weighted consensus networks to represent alternative evolutionary histories.

Mehdi Layeghifard,Vladimir Makarenkov,Pedro R Peres-Neto

doi:10.1186/1471-2148-13-274

Mehdi Layeghifard, Vladimir Makarenkov + Show 1 more

Open Access

https://doi.org/10.1186/1471-2148-13-274

Copy DOI

Abstract

BackgroundThe advent of molecular biology techniques and constant increase in availability of genetic material have triggered the development of many phylogenetic tree inference methods. However, several reticulate evolution processes, such as horizontal gene transfer and hybridization, have been shown to blur the species evolutionary history by causing discordance among phylogenies inferred from different genes.MethodsTo tackle this problem, we hereby describe a new method for inferring and representing alternative (reticulate) evolutionary histories of species as an explicit weighted consensus network which can be constructed from a collection of gene trees with or without prior knowledge of the species phylogeny.ResultsWe provide a way of building a weighted phylogenetic network for each of the following reticulation mechanisms: diploid hybridization, intragenic recombination and complete or partial horizontal gene transfer. We successfully tested our method on some synthetic and real datasets to infer the above-mentioned evolutionary events which may have influenced the evolution of many species.ConclusionsOur weighted consensus network inference method allows one to infer, visualize and validate statistically major conflicting signals induced by the mechanisms of reticulate evolution. The results provided by the new method can be used to represent the inferred conflicting signals by means of explicit and easy-to-interpret phylogenetic networks.

Highlights

The advent of molecular biology techniques and constant increase in availability of genetic material have triggered the development of many phylogenetic tree inference methods
The inference and validation of explicit phylogenetic networks is the main goal of the current study. Another key factor that contributes to the incompatibility among gene trees is stochastic errors resulting from analytical features such as choice of optimality criterion, taxon sampling and sequence evolution model [12,13,14]. These complications makes it difficult for researchers to find reliable estimates of the true species phylogenies, and obstruct such fields as comparative biology and community phylogenetics which rely on phylogenetic trees in their analyses [15,16,17]
We present a new algorithm for the inference of explicit weighted consensus networks from a collection of trees, with or without prior knowledge of the species phylogeny

Summary

Introduction

The advent of molecular biology techniques and constant increase in availability of genetic material have triggered the development of many phylogenetic tree inference methods. Molecular data have played an instrumental, and usually indispensable, role in many phylogenetic and evolutionary studies in the recent decades Their increasing availability is due to outstanding advances in the development of fast, efficient and affordable sequencing technologies [1]. The field of phylogenetics still faces some key analytical challenges stemming from reticulate evolution They include: 1) horizontal gene transfer (e.g., in bacterial or viral evolution); 2) hybridization among species (e.g., allopolyploidy in plants); 3) genetic. Another key factor that contributes to the incompatibility among gene trees is stochastic errors resulting from analytical features such as choice of optimality criterion, taxon sampling and sequence evolution model [12,13,14]. These complications makes it difficult for researchers to find reliable estimates of the true species phylogenies, and obstruct such fields as comparative biology and community phylogenetics which rely on phylogenetic trees in their analyses [15,16,17]

Methods

Results

Conclusion