Coalescent-based species tree inference from gene tree topologies under incomplete lineage sorting by maximum...

Area of Science:

• Phylogenetics
• Computational Biology
• Evolutionary Biology

Background:

• Incomplete lineage sorting (ILS) creates discrepancies between gene and species histories, challenging phylogenetic inference.
• Accurate species tree reconstruction is crucial for understanding evolutionary relationships.

Purpose of the Study:

• To introduce a novel coalescent-based algorithm, STELLS, for maximum likelihood species tree inference.
• To improve the computational efficiency and accuracy of species tree estimation in the presence of ILS.

Main Methods:

• Developed an improved method for calculating gene tree topology probabilities given a species tree, significantly faster than prior methods.
• Implemented a heuristic search algorithm (STELLS) to infer species trees with maximum likelihood from gene tree sets.
• Utilized coalescent theory to model the effects of incomplete lineage sorting.

Main Results:

• The STELLS algorithm demonstrates higher accuracy compared to existing maximum likelihood methods, particularly with noisy gene tree data.
• STELLS offers significant computational efficiency, enabling its application to large biological datasets.
• The algorithm successfully infers species trees by optimizing likelihood through heuristic search.

Conclusions:

• STELLS provides a more accurate and efficient approach to species tree inference under incomplete lineage sorting.
• The developed method addresses a key challenge in phylogenetics, improving the reliability of evolutionary history reconstruction.
• STELLS is a practical tool for analyzing complex evolutionary datasets and is available for download.