From epidemics to pandemics: elucidating the dynamics of Ebola Virus and SARS-CoV-2
The advent of large-scale viral genomic sequencing has provided a rich source of data to explore the dynamics of infectious disease epidemics. In combination with the field of phylodynamics, which allows the inference of unobserved patterns from a relatively small sample of the true diversity of a virus, it has been used to great effect in the past decade. The most notable examples were during the West African Ebola Virus Disease (EVD) epidemic in 2013-2016 and the COVID-19 pandemic, still ongoing at the time of writing. The genomic datasets from these epidemics can be used to explore the evolution and transmission of viruses at different scales, from the effect of within-host evolution, to small-scale transmission networks, and national and international epidemic dynamics. I begin with the national-scale analysis of the dynamics of Ebola virus in Sierra Leone. I developed a phylogeographic analysis in a generalised linear model framework, at two geographical resolutions and in two epochs. I found that the focus of viral movement shifts from the source location in the east, to the capital city in the west. This chapter explores why different locations were important for viral transmission on a national level, and how well the gravity model of infection applies to the spread of Ebola virus in Sierra Leone through time and across different geographical scales. To address some of the issues in modelling a disease like EVD which has a high degree of superspreading, and to explore the impact of local contact networks, I created ABSynthE (Agent Based Synthetic Epidemic). ABSynthE is a flexible agent based model, which simulates an EVD epidemic across the population of Sierra Leone. ABSynthE outputs coalescent phylogenies, which are then used to obtain transmission parameters at each contact level by fitting to results from chapter 1. I found that without any intervention, just under half of the population of Sierra Leone may have been infected, regardless of which district the epidemic began in. There are now well over 8 million genomic sequences of SARS-CoV-2 available for analysis. Within the UK, the sampling is especially dense, allowing detailed epidemiological and phylodynamic analyses to be undertaken. In chapter 3, I explore the origins of the Alpha variant, the first variant of concern, which arose in South East England. I characterise the long ancestral branch, and find that it has a higher evolutionary rate compared to the background and Alpha clades, as well as a single intermediate sequence. I investigate the branches ancestral to the other variants of concern, and explore their mutational profiles, finding that Beta, Gamma and Omicron (but not Delta) have evidence for evolving in a similar manner to Alpha. I explore three different hypotheses for what this manner may be, and conclude that the most likely option is that they evolved within a persistently-infected, but not necessarily immunocompromised, host. Finally, I use the rich UK genomic SARS-CoV-2 dataset to elucidate the dynamics of the first wave of infection in early 2020, including the emergence of the D614G mutation; the Alpha wave, spreading from Kent and London to the rest of the country in late 2020 and early 2021; and the Delta wave, introduced into multiple regions, but mostly spreading from the North West of England in early 2021. I compare these waves, especially in terms of spreading from multiple introductions versus a single origin; and in the context of tightening or loosening non-pharmaceutical interventions.