Presentation Title
Resilience of Viral Genotype Networks Toward Strain-Transcending Immunity
Abstract
Viral genotype networks are real-world examples of spreading processes that evolve across multiple sub-types or strains. Understanding the forces influencing the shape of viral genotype networks, and evolution from strain to strain, may allow genotype network analysis to complement existing predictive modeling of influenza. We compiled the largest network dataset of influenza A hemagglutinin protein sequences to date and found, as previously reported, a network with a highly skewed degree distribution, with central hubs and a low clustering coefficient. We then implemented a multi-strain SIRS model featuring spontaneous mutations and strain-transcending immunity to: (i) explore the role of genotype network structure in determining disease prevalence, and (ii) determine to what extent the observed genotype network is determined by selective pressure due to learned immunity. The influenza HA genotype network was found to be more resistant to increasing strain-transcending immunity than classic toy models such as star-shaped networks, and contains loops that prevent a trapping phenomena observed in chains of nodes. Indeed, if we control for network density, we find that the real influenza network leads to outbreaks more resilient to increases in transcending immunity. Altogether, our results support our hypothesis that observations of viral genotype networks are shaped by strain-transcending immunity to be less centralized and contain loops, to be more similar to small-world networks.
Primary Faculty Mentor Name
Laurent Hébert-Dufresne
Status
Undergraduate
Student College
College of Engineering and Mathematical Sciences
Second Student College
College of Arts and Sciences
Program/Major
Data Science
Second Program/Major
Mathematics
Primary Research Category
Engineering & Physical Sciences
Secondary Research Category
Biological Sciences
Tertiary Research Category
Health Sciences
Resilience of Viral Genotype Networks Toward Strain-Transcending Immunity
Viral genotype networks are real-world examples of spreading processes that evolve across multiple sub-types or strains. Understanding the forces influencing the shape of viral genotype networks, and evolution from strain to strain, may allow genotype network analysis to complement existing predictive modeling of influenza. We compiled the largest network dataset of influenza A hemagglutinin protein sequences to date and found, as previously reported, a network with a highly skewed degree distribution, with central hubs and a low clustering coefficient. We then implemented a multi-strain SIRS model featuring spontaneous mutations and strain-transcending immunity to: (i) explore the role of genotype network structure in determining disease prevalence, and (ii) determine to what extent the observed genotype network is determined by selective pressure due to learned immunity. The influenza HA genotype network was found to be more resistant to increasing strain-transcending immunity than classic toy models such as star-shaped networks, and contains loops that prevent a trapping phenomena observed in chains of nodes. Indeed, if we control for network density, we find that the real influenza network leads to outbreaks more resilient to increases in transcending immunity. Altogether, our results support our hypothesis that observations of viral genotype networks are shaped by strain-transcending immunity to be less centralized and contain loops, to be more similar to small-world networks.