EnsemblCompara GeneTrees: Complete, duplication-aware phylogenetic trees in vertebrates.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Accurate gene family analysis is crucial for understanding evolutionary relationships.
• Existing methods for ortholog prediction and phylogenetic tree generation have limitations, especially for large gene families.

Purpose of the Study:

• To develop a comprehensive, gene-oriented phylogenetic resource called EnsemblCompara GeneTrees.
• To evaluate and compare different phylogenetic tree-building methods.
• To assess the performance of phylogenetic approaches versus clustering for ortholog prediction.

Main Methods:

• Developed a computational pipeline for gene clustering, multiple sequence alignment, and phylogenetic tree generation.
• Created novel, non-sequence-based metrics to assess gene tree correctness.
• Benchmarked various tree-building methods, including TreeBeST from TreeFam.
• Compared phylogenetic tree-based ortholog prediction with traditional clustering methods.

Main Results:

• The TreeBeST method demonstrated superior performance in gene tree reconstruction within this study.
• The phylogenetic approach significantly increased ortholog prediction coverage compared to clustering methods.
• EnsemblCompara GeneTrees provides a robust and scalable resource for gene family analysis.

Conclusions:

• EnsemblCompara GeneTrees offers an improved method for gene family analysis and ortholog prediction.
• The developed resource enhances evolutionary insights by providing comprehensive phylogenetic data.
• All data are publicly available and will be maintained in sync with the Ensembl project.