From viral evolution to spatial contagion: a biologically modulated Hawkes model

Area of Science:

• Epidemiology
• Viral Evolution
• Computational Biology

Background:

• Evolving pathogens can acquire increased contagiousness through mutations.
• Understanding pathogen evolution and geographic spread is crucial for epidemic control.
• Current methods often analyze viral genome data for evolutionary history separately from spatial spread dynamics.

Purpose of the Study:

• To develop a novel model integrating pathogen evolution with spatial contagion dynamics.
• To apply this model to analyze the 2014-2016 Ebola outbreak in West Africa.
• To identify individual viral lineages with significantly high spatiotemporal propagation rates.

Main Methods:

• Proposed a phylogenetic Hawkes process model combining phylogenetic inference and self-exciting process modeling.
• Applied a Bayesian analysis to 23,421 Ebola virus cases from the 2014-2016 outbreak.
• Developed massively parallel implementations for gradient and Hessian calculations for high-performance computing.

Main Results:

• The phylogenetic Hawkes process model successfully integrated viral evolution and spatial spread.
• Identified a subset of 1610 viral samples with associated genome data.
• Detected individual viruses exhibiting significantly elevated rates of spatiotemporal propagation.

Conclusions:

• The developed model effectively links pathogen evolution to spatial contagion dynamics.
• This approach enhances the ability to track and understand epidemic spread.
• High-performance computing frameworks are essential for applying such models to large-scale genomic data.