Tracking Hidden Epidemics: A Phylodynamic Framework for Estimating Pathogen Transmission Dynamics in Locations and Populations with Limited Data Availability

• Applicant: Paula Weidemüller, Ph.D.
• Subject Area: Epidemiology and Medical Biometry/Statistics ; Bioinformatics and Theoretical Biology ; Public Health, Healthcare Research, Social and Occupational Medicine
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 576371952

Effectively responding to infectious disease outbreaks requires timely and accurate estimates of how pathogens spread. Yet, in many parts of the world, including regions where disease burden is highest, surveillance systems are limited. As a result, critical information about transmission dynamics remains hidden, limiting the ability of public health agencies to detect outbreaks early or predict when a pathogen might take hold in a new area. This project aims to address that gap by developing HIDDEN, a new phylodynamic framework that infers transmission dynamics in poorly monitored “source” populations by analyzing pathogen genetic data collected from well-monitored “sink” populations, which experience frequent pathogen introductions from the sources. The project has three main objectives: 1. Model innovation: I will extend the widely used structured birth-death phylodynamic model to integrate information on individual travel history into the pathogen transmission dynamics parameter inference. This approach will allow the model to improve pathogen transmission dynamics within and across locations and populations, increasing the accuracy of estimates even when local data are sparse. 2. Computational acceleration: To ensure the model can be applied in short time frames and to large, complex datasets, I will use neural networks to speed up computationally intensive steps in the parameter inference process. This will allow researchers and public health agencies to perform robust phylodynamic inference in short time frames without requiring high-performance computing resources, allowing them to update the analysis frequently as new data becomes available. These two objectives will be implemented in the BEAST2 ecosystem as a new phylodynamic package called HIDDEN. 3. Application to dengue in the Americas: While many Latin American countries with high dengue virus transmission have limited detection and sequencing infrastructure, regions such as Florida in the U.S. collect high-quality genomic data from local and travel-associated dengue cases. I will apply HIDDEN to thousands of dengue sequences sampled in the U.S. and Latin America. This analysis will map inter-country transmission patterns, reveal under-reported outbreaks, and estimate the risk of dengue becoming endemic in parts of the U.S. The findings will also help identify where and when public health interventions are most needed. Beyond dengue, the HIDDEN framework is designed to be pathogen-agnostic and can be extended to emerging zoonoses like avian influenza, including outbreaks that cross from wild birds into livestock and potentially humans. Through the proposed innovations and analyses, the project contributes to a better understanding of transmission dynamics across multiple regions and provides a valuable foundation for research and public health efforts worldwide and strengthens our ability to detect future epidemics early and respond effectively.

• DFG Programme: WBP Fellowship
• International Connection: USA
• Host: Dr. Nicola Müller