The Occurrence Birth-Death Process for Combined-Evidence Analysis in Macroevolution and Epidemiology

Area of Science:

• Phylogenetics and Evolutionary Biology
• Computational Biology
• Epidemiology

Background:

• Phylodynamic models infer phylogenetic relationships and lineage dynamics from molecular data.
• Estimating past population size (prevalence in epidemiology, paleodiversity in macroevolution) is crucial.
• Total-evidence analyses combine molecular and morphological data, but occurrence data (time-stamped individuals without other data) have been underutilized.

Purpose of the Study:

• To develop new methods for the fossilized birth-death process to incorporate occurrence data.
• To enable piecewise-constant birth, death, and sampling rates estimation.
• To estimate the past number of lineages, with or without an existing tree.

Main Methods:

• Incorporation of occurrence data into the likelihood function of the fossilized birth-death model.
• Implementation of piecewise-constant rates for birth, death, and sampling.
• Development of methods to estimate lineage-through-time, with or without a known phylogeny, within the RevBayes software.

Main Results:

• Successful integration of occurrence data into phylodynamic inference.
• Validation of the method through simulations under various conditions.
• Demonstrated application in inferring COVID-19 prevalence on the Diamond Princess ship and cetacean diversity.

Conclusions:

• The new method enhances phylodynamic inference by leveraging occurrence data.
• This approach bridges traditional epidemiology and pathogen genomics, and paleontology and molecular phylogenetics.
• Joint modeling of occurrences and trees offers a powerful framework for diverse evolutionary and epidemiological studies.