A LIKELIHOOD JUSTIFICATION OF PARSIMONY.

Area of Science:

• Evolutionary biology
• Phylogenetics
• Computational biology

Background:

• Parsimony and likelihood are key methods for inferring evolutionary trees.
• Understanding the relationship between these methods is crucial for robust phylogenetic inference.
• Previous work has highlighted potential statistical inconsistencies in phylogenetic methods.

Purpose of the Study:

• To establish a theoretical connection between maximally parsimonious cladograms and trees of highest likelihood.
• To explore the assumptions underlying this connection within evolutionary theory.
• To discuss the implications for alternative phylogenetic inference methods and statistical consistency.

Main Methods:

• Establishing a theoretical link between cladistics (parsimony) and statistical likelihood approaches.
• Deriving assumptions from the fundamental structure of evolutionary theory.
• Analyzing the independence of these assumptions from the rate of homoplasy.

Main Results:

• A formal connection is demonstrated between maximally parsimonious cladograms and highest likelihood trees.
• The theoretical assumptions underpinning this connection are shown to be independent of homoplasy frequency.
• The findings provide a justification for parsimony methods within a likelihood framework.

Conclusions:

• The study provides a theoretical bridge between parsimony and likelihood in phylogenetics.
• This connection offers a stronger foundation for parsimony-based tree inference.
• The results have implications for understanding the statistical properties of various phylogenetic methods, including potential inconsistencies.