Fast and accurate methods for phylogenomic analyses.

Area of Science:

• Phylogenetics
• Computational Biology
• Evolutionary Biology

Background:

• Species phylogenies are inferred from gene datasets, not directly estimated.
• Gene trees can deviate from species trees due to horizontal gene transfer, incomplete lineage sorting, and gene duplication/loss.
• Various methods exist to estimate species trees from gene trees, but their performance is not well understood.

Purpose of the Study:

• To evaluate the performance of different species tree estimation methods.
• To compare the accuracy of various phylogenetic inference techniques.
• To identify reliable methods for species tree reconstruction.

Main Methods:

• Simulated sequence evolution under a complex model incorporating insertions/deletions (indels), substitutions, and incomplete lineage sorting.
• Evaluation of multiple methods for estimating species trees from simulated sequence datasets.
• Comparison of methods that account for gene tree errors versus those that assume gene trees are correct.

Main Results:

• Fast, simple methods for species tree estimation are nearly as accurate as computationally intensive, sophisticated methods.
• Methods explicitly considering errors in estimated gene trees yield more accurate species trees.
• Accurate species tree estimation is achievable even with gene tree discordance.

Conclusions:

• Highly accurate species tree estimations are possible despite gene tree heterogeneity.
• Computationally tractable and relatively simple methods can achieve high accuracy in species tree reconstruction.
• The study provides insights into the relative performance of different species tree inference methods.