Treeline Provides a Unified Strategy for Optimising Phylogenetic Trees Under Alternative Criteria

Area of Science:

• Phylogenetics
• Computational Biology
• Evolutionary Biology

Background:

• Optimality criteria significantly influence phylogenetic tree construction.
• Existing methods require multiple programs for optimizing trees across different objectives, hindering comparative analyses.
• Recent findings question the assumption that computationally intensive objectives always yield superior phylogenetic trees.

Purpose of the Study:

• To introduce Treeline, a unified software for optimizing phylogenetic trees under balanced minimum evolution, maximum likelihood, and maximum parsimony objectives.
• To facilitate direct comparison of phylogenetic trees across various optimization objectives and evolutionary models.
• To simplify the process of constructing and comparing phylogenetic trees.

Main Methods:

• Treeline employs a novel optimization strategy by perturbing the patristic distance matrix for initializing candidate trees.
• The software integrates multiple optimization objectives within a single framework.
• Performance was evaluated against existing single-objective programs in terms of accuracy, memory efficiency, and speed.

Main Results:

• Treeline demonstrated comparable or superior accuracy, memory efficiency, and speed compared to single-objective programs.
• The balanced minimum evolution objective yielded gene trees more consistent with species trees than maximum likelihood or maximum parsimony.
• A case study on Hominidae species illustrated Treeline's utility in constructing trees under alternative objectives and models.

Conclusions:

• Treeline offers a unified and efficient approach to phylogenetic tree optimization and comparison.
• The balanced minimum evolution objective shows promise for accurate species tree inference.
• Treeline simplifies comparative phylogenetic analyses, aiding evolutionary research.