The performance of outgroup-free rooting under evolutionary radiations

Area of Science:

• Phylogenetics
• Evolutionary Biology
• Computational Biology

Background:

• Phylogenetic tree rooting is crucial for understanding ancestral-descendant relationships, requiring root node placement.
• Traditional rooting relies on outgroup lineages, which are not always available or identifiable.
• Outgroup-free methods, such as midpoint rooting (MPR) and minimal ancestor deviation (MAD), offer alternatives but vary in performance.

Purpose of the Study:

• To evaluate the performance of MPR and MAD methods in rooting phylogenetic trees under challenging evolutionary conditions.
• To identify factors influencing the success rates of these outgroup-free rooting techniques.
• To develop a predictive model for rooting accuracy in specific tree topologies.

Main Methods:

• Analysis of MPR and MAD performance on simulated phylogenetic trees representing fast evolutionary radiations with short internal branches.
• Investigation of tree balance impact on the relative success of MPR and MAD.
• Development of a general linear model using dimensionless metrics (coefficient of variation of external branch lengths, ratio of internal to total branch length) to predict rooting success for four-taxa trees.

Main Results:

• Both MPR and MAD exhibited average success rates below 50% in challenging scenarios, with MAD slightly outperforming MPR.
• Tree balance significantly affected the relative performance of the two methods.
• A predictive model based on branch length metrics was devised, showing potential for estimating rooting accuracy.

Conclusions:

• Outgroup-free rooting methods face limitations in accurately rooting phylogenies from rapid radiations.
• Predictive metrics can offer insights into the reliability of rooting attempts for specific tree structures.
• The developed model predicts moderate to good performance (50-75%) for these methods in empirical datasets like placental mammal radiation.