A priori vs emergent low-dimensionality of antigenic spaces

Ascertain whether the empirically reconstructed antigenic spaces of pathogen strains are intrinsically low-dimensional or whether their apparent low dimensionality emerges from epidemiological and evolutionary dynamics or sampling effects, in order to clarify the foundations of low-dimensional antigenic mapping used in infectious disease modeling.

Background

The paper develops an evo-SIS framework coupling pathogen transmission on a metapopulation network with mutation on a strain network, using diffusion geometry to encode cross-immunity. Many prior models effectively assume low-dimensional antigenic spaces, motivated by empirical mappings for pathogens such as influenza and SARS-CoV-2.

The authors emphasize that while low-dimensional antigenic maps are often observed, it remains unclear whether this structure is fundamental to antigenic relations or instead arises from epidemiological-evolutionary feedbacks or sampling biases. Resolving this uncertainty would ground the use of low-dimensional projections in a mechanistic understanding.