Genomic Epidemiology across Scales: From SARS-CoV-2 Intrahost Diversity to National Trends in Group B Streptococcus Invasive Disease

Abstract

Pathogen virulence factors and evolutionary processes shape the clinical outcomes and epidemiological dynamics of infectious diseases. Therefore, understanding these is essential for developing and ensuring the sustained effectiveness of prevention, treatment, and surveillance strategies. In this dissertation, I present genomic epidemiology studies across complementary scales of pathogen diversity (from intrahost to population-level) and human populations (from local to national). This work aimed to characterize the pathogen risk factors and evolutionary dynamics underlying two clinically important infectious diseases: SARS-CoV-2 and Group B Streptococcus (GBS). In the first study (Chapter 2), we examine how SARS-CoV-2 evolves within host during acute infections, characterizing the selective pressures acting on intrahost variants, and their transmission dynamics. We find that intrahost diversity in acute infections is limited, subject to purifying selection, and rarely transmitted, making it an unlikely source of novel Variants of Concern (VoCs) and an insensitive marker for inferring transmission events, though it may nonetheless influence clinical outcomes. In the second and third studies (Chapters 3 and 4), we investigate the population-level genetic characteristics of bacterial strains causing neonatal invasive GBS disease at the local and national scales in the United States. At the local scale, we characterize the bacterial features associated with clinical severity using a dataset from Boston Children's Hospital (Chapters 3). We find that the population diversity in this local settings reflect global observations, with comparable circulating strains, antimicrobial resistance trends and estimated vaccine coverage. Although we identify no bacterial features associated with disease severity, we find evidence that a pilus subvariant is associated with age of onset among late- and very-late-onset cases, potentially reflecting evolving susceptibility or exposure to specific antigens over time. Scaling up to a national genomic dataset (Chapters 4), we investigate bacterial determinants of disease onset timing, with the aim of uncovering the acquisition source of late- and very-late-onset etiologies which are not prevented by current antibiotic prophylaxis. Through a genome-wide association study (GWAS), we find evidence for associations between age of onset and several variants linked to capsular polysaccharide biosynthesis. However, pervasive linkage disequilibrium around this genomic region likely reflects redundant biological signals and prevents the disentanglement of causal variants from hitchhiking ones. Together, these studies highlight the influence of pathogen standing diversity and further diversification on clinical and epidemiological features of infectious diseases, and demonstrate that investigating these processes across multiple pathogen and human population scales can yield complementary and valuable insights.