Phylogeny-based estimates of epidemic population substructure outperform time series approaches

Improving infectious disease models plays a crucial role in mitigating global health impacts. Population-level models have been essential for understanding and anticipating impacts of infectious diseases, but their accuracy has often been limited due to changing environments, rapid pathogen evolution, and insufficient access to detailed transmission data. In particular, due to the lack of granular transmission data, many epidemiological models assume homogeneous disease transmission rather than acknowledge differences between groups within the host population. To address this gap, we conducted a simulation study to assess the potential for inferring group-structured viral transmission dynamics from phylogenetic data and to demonstrate how age structure can be incorporated in case projections. Using a synthetic dataset of 800 age-structured viral phylogenies and their associated case time series, we estimated transmission rates within and between age groups. Our results show that the estimates derived from phylogenies are more accurate than those from case counts alone, although both approaches enabled recovery of age-specific transmission differences. These results suggest that viral phylogenies and case-count time series each provide valuable data that can improve our understanding of population-level disease dynamics, ultimately contributing to more accurate and effective models for informing public policy and reducing disease burden.