Estimating Epidemic Incidence and Prevalence from Genomic Data.

Area of Science:

• Epidemiology
• Computational Biology
• Genomics

Background:

• Phylodynamic methods model epidemic transmission using phylogenetic trees.
• Current models struggle to explicitly infer past incidence and prevalence.
• Birth-death and coalescent processes are used in epidemiological models like SIS and SIR.

Purpose of the Study:

• To develop a novel particle filtering framework for explicit inference of epidemic trajectories.
• To integrate genomic sequences and case count data for enhanced phylodynamic analysis.
• To apply the method to real-world epidemic data, such as the Ebola outbreak.

Main Methods:

• A particle filtering framework is introduced to model birth-death processes.
• The method explicitly infers prevalence and incidence trajectories alongside phylogenies.
• It utilizes genomic sequences and epidemiological case count data.

Main Results:

• The particle filtering framework demonstrates accuracy on simulated data.
• The method successfully infers past incidence and prevalence trajectories.
• Application to the 2014 Ebola outbreak in Sierra Leone provides insights into disease dynamics.

Conclusions:

• The developed particle filtering method offers a powerful tool for phylodynamic inference.
• Explicitly inferring incidence and prevalence enhances our understanding of epidemic dynamics.
• This approach has significant implications for real-time epidemic monitoring and response.