ASTRAL-Pro: Quartet-Based Species-Tree Inference despite Paralogy.

Area of Science:

• Evolutionary biology
• Genomics
• Bioinformatics

Background:

• Phylogenomics uses genome-wide data to infer evolutionary histories, accounting for conflicting gene trees.
• Current summary methods like ASTRAL excel at inferring species trees from single-copy genes but ignore gene duplication and loss events.
• This limitation leads to discarding substantial genomic data in phylogenomic studies.

Purpose of the Study:

• To develop a novel method for accurate species tree inference that incorporates both single-copy and multicopy gene trees.
• To address the limitations of existing phylogenomic methods that exclude data from gene duplications and losses.
• To improve the accuracy and scalability of phylogenetic analyses using comprehensive genomic data.

Main Methods:

• Proposed a quartet similarity measure to compare single-copy and multicopy gene trees, accounting for orthology and paralogy.
• Introduced ASTRAL-Pro, a dynamic programming method to optimize this quartet similarity measure for species tree reconstruction.
• Evaluated ASTRAL-Pro's performance using extensive simulated and real biological datasets.

Main Results:

• ASTRAL-Pro demonstrates superior accuracy in species tree inference compared to existing alternative methods.
• The new method effectively utilizes information from both single-copy and multicopy genes, including those affected by duplication and loss.
• Performance evaluations on simulated and real data confirm the robustness and accuracy of ASTRAL-Pro.

Conclusions:

• ASTRAL-Pro represents a significant advancement in phylogenomics, enabling the analysis of previously underutilized multicopy gene data.
• The method enhances the accuracy of species tree reconstruction by modeling gene duplication and loss.
• This work facilitates more comprehensive and accurate evolutionary studies by leveraging the full spectrum of genomic information.