Quartet inference from SNP data under the coalescent model.

Area of Science:

• Phylogenetics
• Computational Biology
• Evolutionary Biology

Background:

• Estimating species-level phylogenetic relationships under the coalescent model is crucial but challenging.
• Existing methods often rely on summary statistics (gene trees) or computationally intensive algorithms, limiting scalability to large genomic datasets.

Purpose of the Study:

• To develop a computationally efficient method for inferring species phylogenetic relationships using algebraic statistics.
• To quantify uncertainty in phylogenetic estimates using the nonparametric bootstrap.
• To demonstrate the method's utility for species tree inference in large datasets.

Main Methods:

• Inference of phylogenetic relationships among quartets of taxa using techniques from algebraic statistics.
• Quantification of uncertainty via nonparametric bootstrap.
• Implementation in the SVDquartets software.

Main Results:

• The developed method accurately infers phylogenetic relationships among quartets under the coalescent model.
• Nonparametric bootstrap effectively quantifies uncertainty in estimated relationships.
• The method demonstrates utility for species tree inference in empirical datasets (rattlesnakes, soybeans).

Conclusions:

• The novel algebraic statistics-based method provides an efficient and scalable approach for species phylogenetic inference.
• This method addresses limitations of existing techniques, particularly for whole-genome data.
• SVDquartets software facilitates the application of this method in evolutionary studies.