Contagion Spread and Networks

Abstracts

Colin Carlson

Cross-species contagion: viral macroecology in a shifting biosphere

Conventional wisdom says pathogens spread between hosts on ecological timescales, and between host species over evolutionary timescales. However, within roughly a century, our world has suddenly warmed 1.3 degrees, lost hundreds of millions of hectares of intact rainforest, and gained billions of livestock. As a result, pathogens are spreading among wildlife, vectors, livestock, and humans at an unprecedented speed. In my talk, I will discuss how machine learning models can be used to predict this spread: first, by assessing compatibility between hosts and pathogens; and second, by anticipating opportunities for cross-species transmission. I will conclude by discussing the challenges of bridging the host and host species scale with multi-layer contact networks, and opportunities to continue integrating network approaches into viral macroecology and pandemic prevention.

Jacob Liam Curran-Sebastian

Transmission Networks and Intervention Effects from SARS-CoV-2 Genomic and Social Network Data in Denmark

The COVID-19 pandemic saw governments increasingly making use of large-scale pathogen genomics for decision making. We use 293,841 SARS-CoV-2 genomes collected in Denmark between September 1st 2020 and December 31st 2021 and combine these with comprehensive individual-level data on social settings, including households, schools, workplaces, and family relationships. We develop scalable tools to infer a network of plausible transmission pairs from this data. From this network we sample plausible transmission trees and identify over 7,000 transmission clusters associated with these settings. We further investigate the effectiveness of specific non-pharmaceutical interventions (NPIs) and quantify transmission heterogeneities, providing a more detailed understanding of disease spread than is possible from aggregate national estimates. Our approach is pathogen agnostic and can be used in future outbreaks where genomic data and data on social relationships are available.

Leah Keating

Loops, not groups: Long cycles are responsible for discontinuous phase transitions in higher-order contagions

Discontinuous phase transitions are often observed in the outbreak sizes when we have dynamics on higher-order networks. Here, we consider higher-order networks to be networks with groups. In this talk, we consider complex-contagion dynamics on a network with groups of size two and three. We show that just having groups is insufficient to observe a discontinuous phase transition in the outbreak size, and that longer cycles are required to produce this behaviour. This brings us closer to fully understanding why we sometimes see discontinuous phase transitions in dynamics on higher-order networks where we do not in the dyadic versions of the models.

Juliana Taube

Putting network epidemiology theory to the test: Estimating contact structure across epidemic conditions with implications for control

The field of network epidemiology has deepened our understanding of how human behavior, especially contact heterogeneity, structures infectious disease transmission. Further advances to epidemic prediction and control are limited by a lack of empirical data on contact structure and how contact structure is affected by disease transmission. We address these gaps using large-scale contact survey data from the US over eleven months during the COVID-19 pandemic. Using spatiotemporal GAMs and INLA techniques, we find substantial individual and spatial heterogeneity in contact patterns, with changes in average contact rates highly associated with fluctuations in state and national case incidence. By examining contact patterns disaggregated by infection status, we validate theoretical predictions about the relationship between node degree and time of infection. Finally, we address critical questions about how non-pharmaceutical interventions affect contact heterogeneity and herd immunity thresholds, with implications for policy implementation and exit in future epidemic responses. Together, this work provides detailed data that can be used to parameterize future models, key evidence in support of coupled disease-behavior modeling assumptions, and reframes how we think about contact heterogeneity in the context of epidemic interventions.

Nicholas Landry

Reconstructing networks from simple and complex contagions

Network scientists often use complex dynamic processes to describe network contagions, but tools for fitting contagion models typically assume simple dynamics. I'll address this gap by presenting a nonparametric method to reconstruct a network and dynamics from a series of node states, using a model that breaks the dichotomy between simple pairwise and complex neighborhood-based contagions. I'll then show that a network is more easily reconstructed when observed through the lens of complex contagions if it is dense or the dynamic saturates, and that simple contagions are better otherwise. I'll conclude by talking about the inadequacies of the term "complex contagion".

Laurent Hebert-Dufresne

Simpson's contagions

Complex contagions describe systems where the probability or rate of contagious transmission is a nonlinear function of the exposure to contagious agents. Recent studies have shown that local correlations (e.g., group structure or temporal burstiness) and heterogeneity (e.g., diversity of parameters or covariates) can give the illusion of nonlinear effects even when the dynamics is actually linear. We briefly review these studies to inform a new model and explanation for these effective models of complex contagions. We find global threshold dynamics and superlinear complex contagions even in populations where agents are distributed across social groups described solely by linear or even sublinear contagions. This effect can be understood as a manifestation of Simpson's paradox. Incidence data from heterogeneous groups can look superlinear once averaged over all groups, since the sampling of groups represented at high incidence is biased towards those with stronger local transmission. We then define what we call a Simpson's contagion: a contagion process that looks superlinear when observed over an entire population, but is mechanistically linear or even sublinear in all of its subgroups.