What if I don't have a tree? Split decomposition and related models.

Area of Science:

• Evolutionary biology
• Bioinformatics
• Computational phylogenetics

Background:

• Phylogenetic analyses often yield conflicting signals from sequence or distance data, hindering the determination of a single evolutionary tree.
• Traditional phylogenetic methods may struggle to represent complex evolutionary histories with reticulate evolution or incomplete lineage sorting.

Purpose of the Study:

• To introduce and describe the SplitsTree program, a software tool for phylogenetic network construction.
• To present the split decomposition method as a robust approach for analyzing data with conflicting phylogenetic signals.

Main Methods:

• The study implements the split decomposition method to analyze phylogenetic data.
• SplitsTree computes and visualizes phylogenetic networks, referred to as splits graphs.
• The program incorporates various distance transformations, parsimony splits, spectral analysis, and bootstrapping.

Main Results:

• Split decomposition can resolve conflicting phylogenetic signals by representing data as a tree-like network.
• For ideal data, the method produces a standard phylogenetic tree.
• For less ideal data, it generates networks that reveal evidence of multiple, potentially conflicting, evolutionary histories.

Conclusions:

• The SplitsTree program provides a powerful tool for visualizing and analyzing complex phylogenetic relationships beyond simple tree structures.
• Split decomposition effectively handles datasets with ambiguous or conflicting evolutionary signals.
• Phylogenetic networks generated by SplitsTree offer a more nuanced representation of evolutionary history.