Distinguishing Phylogenetic Level-2 Networks with Quartets and Inter-Taxon Quartet Distances

Area of Science:

• Evolutionary Biology
• Phylogenetics
• Bioinformatics

Background:

• Phylogenetic networks model complex evolutionary events like hybridization and gene flow.
• Current inference methods are limited to simpler phylogenetic level-1 networks.
• Statistically consistent inference for more complex networks is a significant challenge.

Purpose of the Study:

• To establish theoretical foundations for statistically consistent inference of semi-directed level-2 phylogenetic networks.
• To broaden the scope of phylogenetic network inference beyond level-1 networks.
• To enable robust inference from quartet data.

Main Methods:

• Characterizing distinguishable features between level-2 networks and their displayed quartet trees.
• Proving circular decomposability of inter-taxon distances derived from quartets.
• Developing a framework for inference from quartet data under the Network Multispecies Coalescent model.

Main Results:

• Theoretical foundations for statistically consistent inference of semi-directed, outer-labeled, planar, galled level-2 networks are established.
• Features distinguishing these networks from their displayed quartet trees are precisely characterized.
• Circular decomposability of inter-taxon distances from quartets is proven, enabling robust inference.

Conclusions:

• The study provides a significant advancement in phylogenetic network inference.
• The developed methods allow for more accurate modeling of complex evolutionary histories.
• Novel identifiability implications are presented for various data types and evolutionary models.