Phylogeographic Inference of SARS-CoV-2 Delta Wave in Texas, USA using a Novel Spatial Transmission Count Statistic (preprint)

Viral genomes contain records of geographic movements and cross-scale transmission dynamics. However, the impact of population heterogeneity, particularly among rural and urban areas, on viral spread and epidemic trajectory has been less explored due to limited data availability. Intensive and widespread efforts to collect and sequence SARS-CoV-2 viral samples have enabled the development of comparative genomic approaches to reconstruct spatial transmission history and understand viral transmission across different scales. Large genomic datasets with few mutations present challenges for traditional phylodynamic approaches. To address this issue, we propose a novel spatial transmission count statistic that efficiently summarizes the geographic transmission patterns imprinted on viral phylogenies. Our analysis pipeline reconstructs a time-scaled phylogeny with ancestral trait states and identifies spatial transmission linkages, categorized as imports, local transmission, and exports. These linkages are summarized to represent the epidemic profile of the focal area. We demonstrate the utility of this approach for near real-time outbreak analysis using over 12,000 full genomes and linked epidemiological data to investigate the spread of the SARS-CoV-2 Delta variant in Texas. Our goal is to trace the Delta variants origin, timing and to understand the role of urban and rural areas in the spatial diffusion patterns observed in Texas. Our study shows (1) highly populated urban centers were the main sources of the epidemic in Texas; (2) the outbreaks in urban centers were connected to the global epidemic; and (3) outbreaks in urban centers were locally maintained, while epidemics in rural areas were driven by repeated introductions. Significance StatementWe developed a novel phylogeographic approach that analyzes transmission patterns at low computational cost. This method not only facilitates the inference of spatial scales of transmission but also enables exploration of how specific demographic characteristics influence transmission patterns among heterogenous populations. The rural population in the US, comprising approximately 60 million individuals, has been significantly impacted by COVID-19. Applying our new method, we examined the variations in epidemic patterns between urban centers (e.g., Houston) and rural areas in Texas. We found that urban centers are the primary source for SARS-CoV-2 in rural areas. This analysis lays the groundwork for designing effective public health interventions specifically tailored to the needs of affected areas.

Phylogeny and Metadata Network Database for Epidemiologic Surveillance (preprint)

The ongoing SARS-CoV-2 pandemic has highlighted the difficulty in integrating disparate data sources for epidemiologic surveillance. To address this challenge, we have created a graph database to integrate phylogenetic trees, associated metadata, and community surveillance data for phylodynamic inference. As an example use case, we divided 22,713 SARS-CoV-2 samples into 5 groups, generated maximum likelihood trees, and inferred a potential transmission network from a forest of minimum spanning trees built on patristic distances between samples. We then used Cytoscape to visualize the resultant graphs.