Exploring power and parameter estimation of the BiSSE method for analyzing species diversification

Area of Science:

• Evolutionary Biology
• Phylogenetics
• Computational Biology

Background:

• The BiSSE (Binary State Speciation and Extinction) method is increasingly used to study diversification and character evolution.
• Understanding BiSSE's limitations is crucial for accurate evolutionary inference.

Purpose of the Study:

• To evaluate the Binary State Speciation and Extinction (BiSSE) method's hypothesis testing capabilities.
• To investigate BiSSE's susceptibility to confounding effects from sample size and rate asymmetries.

Main Methods:

• Simulations were conducted using varying sample sizes (number of tips).
• Simulations incorporated different rate asymmetries in speciation, extinction, and character change.
• The impact of high tip ratio bias was assessed.

Main Results:

• BiSSE's statistical power is significantly reduced by small sample sizes and high tip ratio bias.
• Parameter estimation accuracy and precision decreased under these conditions.
• BiSSE struggled to identify the specific rate asymmetry causing an excess of a character state when biased.

Conclusions:

• Caution is advised when interpreting BiSSE results with fewer than 300 terminals.
• High tip ratio bias (less than 10% of species in one state) also warrants careful interpretation.
• BiSSE accurately infers rate asymmetries in low bias scenarios with sufficient sample size.