Fig. 6: Modelling IgG levels following exposure to antigenic proteins.

a Parameter estimates from linear mixed-effects models including data on IgG levels in the maternal blood sample and individuals’ history of exposure to proteins. The left column shows estimates from modelling the IgG response to proteins to which only a subset of individuals were exposed in the first year of life (n = 75). The right column shows estimates from a model of IgG binding to proteins to which all individuals were exposed in the first year of life (n = 85). The top row shows the increase in IgG binding following an individual being exposed to an antigen, which could only be estimated in the partial exposure model. The middle row shows the strength of the relationship between the IgG levels in the maternal birth sample and those at later timepoints. The bottom row shows the expected IgG response to ABTs at 24 mo, following adjustment for other factors included in the model. b Line plots showing the relationship between the number of pneumococcal samples isolated from an individual that encoded a protein, and the median IgG binding to the protein across individuals with the corresponding level of exposure at the time serum was sampled. The pale lines show the median IgG binding level to all ABTs (n = 210). The darker lines highlight the data for proteins to which IgG responses were inferred to at least double after one or more exposures in (a) (n = 8). Source data are provided as a Source Data file.