Theoretical foundation of the minimum-evolution method of phylogenetic inference

Area of Science:

• Evolutionary biology
• Phylogenetics
• Computational biology

Background:

• The minimum-evolution (ME) method is a common approach for inferring phylogenetic trees.
• The assumption that the tree with the minimum sum of branch lengths is the most accurate has lacked rigorous mathematical proof.
• Previous phylogenetic analyses have relied on this unproven assumption.

Purpose of the Study:

• To provide a theoretical foundation for the minimum-evolution (ME) method in phylogenetic inference.
• To mathematically demonstrate that the ME tree has the smallest expected sum of branch lengths.
• To develop formulas for calculating branch lengths and their standard errors using ordinary least-squares.

Main Methods:

• Mathematical derivation of the theoretical basis for the ME method.
• Application of ordinary least-squares (OLS) for estimating branch lengths and standard errors.
• Analysis of mitochondrial DNA (mtDNA) sequence data from humans and a chimpanzee outgroup.

Main Results:

• The study establishes the mathematical proof for the ME method's principle.
• Formulas for computing branch length estimates and standard errors for unrooted bifurcating trees were derived.
• The ME tree for the analyzed mtDNA data showed slight differences from the neighbor-joining tree, but these were not statistically significant.

Conclusions:

• The minimum-evolution method is theoretically sound when using statistically unbiased evolutionary distances and OLS estimation.
• The derived formulas provide a robust framework for phylogenetic tree construction using the ME approach.
• The study validates the ME method's utility in analyzing genetic sequence data for evolutionary insights.