A multi-scale modeling approach for infectious disease: Integrating within-host dynamics to network-based transmission

Infectious disease transmission is shaped by tightly coupled biological processes within hosts and social contact patterns between hosts, making cross-scale modeling essential for understanding epidemic dynamics. We develop a data-driven multiscale framework that links within-host viral kinetics to between-host transmission on a dynamic multilayer contact network. Using influenza virus, SARS-CoV-2, and SARS-CoV as representative pathogens, we infer infection-age-dependent infectiousness from longitudinal viral-load data and integrate the resulting profiles into a stochastic individual-based model with contact layers constructed from empirical demographic and contact data. By comparing infectiousness profiles within the same multilayer transmission framework, we show that epidemic trajectories depend not only on overall transmission potential but also on the timing and duration of infectiousness. Early, concentrated infectiousness produces earlier epidemic peaks and shorter outbreaks, whereas later or prolonged infectiousness delays and extends transmission. Layer-resolved analyses show that setting-specific contributions arise from interactions among contact volume, contact intensity, and infectiousness profiles. Intervention simulations highlight the importance of temporal alignment: case isolation and vaccination are most effective when implemented before or during the pathogen-specific period of high infectiousness, whereas physical distancing reduces epidemic burden by lowering contact opportunities. This framework provides a flexible approach for linking viral-load-derived infectiousness with network-mediated transmission and evaluating targeted epidemic control strategies.