Correlates of Nonrandom Patterns of Serotype Switching in Pneumococcus

[[[ Background: ]]] Pneumococcus is a diverse pathogen, with >90 serotypes, each of which has a distinct polysaccharide capsule. Pneumococci can switch capsules, evading vaccine pressure. Certain serotype pairs are more likely to occur on the same genetic background as a results of serotype switching, but the drivers of these patterns are not well understood. [[[ Methods: ]]] We used the PubMLST and Global Pneumococcal Sequencing Project databases to quantify the number of genetic lineages on which different serotype pairs occur together. We also quantified the genetic diversity of each serotype. Regression model were used to evaluate the relationship between shared polysaccharide components and the frequency of serotype co-occurrence and diversity. [[[ Results: ]]] A number of serotype pairs occurred together on the same genetic lineage more commonly than expected. Co-occurrence of between-serogroup pairs was more common when both serotypes had glucose as a component of the capsule (and, potentially, glucuronic acid, any-N-acetylated sugar, or ribitol). Diversity also varied markedly by serotype and was associated with the presence of specific sugars in the capsule. [[[ Conclusions: ]]] Certain pairs of serotypes are more likely to co-occur on the same genetic background. These patterns were correlated with shared polysaccharide components. This might reflect adaptation of strains to produce capsules with specific characteristics.