Serodynamics: a review of methods for epidemiological inference using serological data

The availability and diversity of serological data measuring antibody responses to infectious pathogens, accelerated in response to the SARS-CoV-2 pandemic, has enabled key insights into infectious disease dynamics and population health. Here, we present a review of analytical approaches and considerations for inference using serological data, highlighting the range of epidemiological and biological insights that are possible using appropriate mathematical and statistical models. This in-depth review focuses on methods to understand transmission dynamics and infer past exposures from serological data, referred to as serodynamics, though we note that such analyses often address complementary immunological questions. We first discuss key considerations for data processing and interpretation of raw serological data which are prerequisite for fitting serodynamical models. We then review a range of approaches for estimating epidemiological trends, ranging from classical serocatalytic models applied to binary serostatus data, to contemporary methods using full quantitative antibody measurements and immunological understanding to estimate if and when individuals have been previously infected. Here, we collate and synthesize these approaches within the context of a unifying framework for the overall data-generation process, consisting of key concepts including antibody kinetics, quantitative models to represent within-host and epidemic processes, and considerations for linking observed serological data to models. We close with a discussion of the types of methodological developments needed to meet the increasingly complex serological data becoming available that provide new avenues for scientific discovery and public health insights.