Serosolver

Serosolver infers individual infection histories, antibody dynamics, and epidemiological parameters from serological datasets to disentangle latent infections and cross-reactive antibody responses (e.g., influenza).

Key Features:

• Dynamic Modeling Framework: Employs a dynamic mechanistic model to infer individual-level infection histories, historical attack rates, and patterns of antibody dynamics.
• Inference of Latent Processes: Jointly infers unobserved prior infections and the parameters governing antibody responses from serological titers.
• Cross-Reactive Antibody Dynamics: Accounts for cross-reactive antibody responses between multiple strains to improve interpretation of serological data.
• Measurement Error Consideration: Incorporates measurement error in serological assay observations during inference.
• Flexibility Across Datasets: Applicable to datasets studying various viruses over different timescales to recover epidemiological and immunological insights.

Scientific Applications:

• Epidemiological Insights: Infers historical attack rates and age-stratified infection risks to characterize population-level transmission patterns.
• Immunological Parameters: Estimates antibody titre boosting, waning, and cross-reaction parameters relevant to immune response and vaccine studies.

Methodology:

Jointly infers latent infection dynamics and antibody responses using a mechanistic model of antibody titers over time linked to serological assay observations, while accounting for measurement error.