Measuring Asymmetry in Time-Stamped Phylogenies

Area of Science:

• Evolutionary biology
• Virology
• Computational biology

Background:

• Phylogenetic asymmetry can indicate transmission heterogeneity in viral populations, such as during acute HIV infection or within core groups.
• Existing statistical tests for phylogenetic asymmetry are prone to false positives with uneven sampling times and false negatives for localized asymmetry.

Purpose of the Study:

• To develop a robust statistical framework for testing phylogenetic asymmetry.
• To identify localized asymmetry within viral phylogenies.
• To provide deeper insights into viral population structures and transmission dynamics.

Main Methods:

• A novel permutation-based approach was developed to test for phylogenetic asymmetry, comparing observed phylogenies against randomized ones with matched sampling and coalescence times.
• Profiles of asymmetry measures across evolutionary timescales were generated to detect localized asymmetry.
• The approach was validated using various asymmetry metrics on real viral datasets.

Main Results:

• The permutation test effectively reduces false positive rates associated with varying sampling times.
• The profile-based method successfully identifies localized asymmetry within phylogenies.
• The combined approach offers a more nuanced understanding of deviations from standard coalescent and birth-death models.

Conclusions:

• The new methods provide more accurate and detailed analyses of viral phylogenetic asymmetry.
• These advancements enhance our ability to infer population structure and transmission heterogeneity from viral genetic data.
• The findings challenge simplistic assumptions in common phylogenetic modeling for viral evolution.