A Metric on Phylogenetic Tree Shapes

Area of Science:

• Phylogenetics
• Computational Biology
• Evolutionary Biology

Background:

• Comparing evolutionary tree shapes is challenging due to difficulties in complete shape description.
• Existing methods lack a standardized approach for quantifying and comparing tree shapes across different evolutionary processes.

Purpose of the Study:

• To present a comprehensive characterization of rooted branching tree shapes for natural comparisons.
• To define a novel metric for quantifying differences in tree shapes.
• To enable more robust comparisons of evolutionary histories and epidemiological patterns.

Main Methods:

• Developed a full characterization of rooted branching tree shapes.
• Defined a distance metric based on this characterization.
• Applied the metric to distinguish between random models and empirical data (Influenza A sequences).
• Extended the metric to incorporate branch lengths and tree imbalance.

Main Results:

• The proposed metric successfully distinguishes trees generated by different random models.
• The metric differentiates evolutionary trees from tropical versus USA Influenza A sequences, reflecting epidemiological differences.
• The characterization allows for tree addition, multiplication, and computation of average tree shapes.

Conclusions:

• The new characterization and metric provide a powerful tool for analyzing and comparing evolutionary tree shapes.
• This approach enhances our ability to study evolutionary processes and understand disease epidemiology.
• The framework supports advanced analyses, including the computation of average tree shapes.