Multi-rate Poisson tree processes for single-locus species delimitation under maximum likelihood and Markov chain...

Area of Science:

• Biodiversity research
• Molecular biology
• Computational phylogenetics

Background:

• Molecular species delimitation is crucial for biodiversity assessment.
• Distance-based methods are common but have limitations (threshold sensitivity, ignoring evolutionary history).
• The Poisson Tree Processes (PTP) method is phylogeny-aware but struggles with divergent intraspecific variation and speed on large datasets.

Purpose of the Study:

• To introduce an improved phylogeny-aware method for molecular species delimitation.
• To address the limitations of existing methods, particularly for large-scale biodiversity surveys.
• To enhance accuracy and computational efficiency in species delimitation.

Main Methods:

• Development of the multi-rate Poisson Tree Processes (mPTP) method.
• Incorporation of varying levels of intraspecific genetic diversity.
• Implementation of a novel dynamic programming algorithm and Markov Chain Monte Carlo (MCMC) sampling.

Main Results:

• mPTP demonstrates superior accuracy in species delimitation compared to PTP and distance-based methods.
• The method consistently identifies more taxonomic species and infers species numbers closer to established taxonomy.
• mPTP achieves a significant speedup (at least five orders of magnitude) for large datasets.
• MCMC sampling provides rapid and comprehensive evaluation of delimitation results.

Conclusions:

• mPTP offers a more accurate and practical approach to molecular species delimitation, especially for large datasets.
• The method overcomes key limitations of previous phylogeny-aware and distance-based techniques.
• mPTP facilitates robust biodiversity estimates and species discovery in molecular surveys.