Phylogeographic Estimation and Simulation of Global Diffusive Dispersal

Area of Science:

• Phylogenetics
• Computational Biology
• Biogeography

Background:

• Estimating dispersal rates is crucial for understanding species invasions and disease spread.
• Current methods often use planar Brownian Motion (BM), which is inaccurate at global scales.
• Spherical Brownian Motion (SBM) better models global dispersal but lacks statistical estimation and simulation tools.

Purpose of the Study:

• To develop statistical methods for simulating SBM along phylogenetic trees (timetrees).
• To create methods for estimating SBM diffusion coefficients from timetrees and geographic data.
• To enable accurate analysis of global dispersal patterns.

Main Methods:

• Developed new methods for simulating SBM on timetrees.
• Adapted Felsenstein's independent contrasts and maximum likelihood for SBM diffusivity estimation.
• Implemented methods in the R package "castor" for efficient computation.

Main Results:

• Successfully simulated SBM along given timetrees.
• Enabled estimation of SBM diffusion coefficients from timetrees and tip coordinates.
• Demonstrated methods on Cyanobacteria and Influenza B virus data.

Conclusions:

• New methods accurately model global dispersal using SBM.
• The "castor" R package provides tools for phylogenetic diffusion analysis.
• Enables new insights into global biogeography and epidemiology.