Speaker
Description
Spillover of infectious diseases is a significant issue for human, animal, and plant health, in part due to their potential to establish in the new host population, thus achieving a host jump. Current approaches to predict and prevent host jumps are structured around discovering new viruses, drawing associations between pathogen characteristics and host jump risk, and managing spillover, but prediction and prevention are still elusive. Spillover is a critical step in the process of completing a host jump, so it is often assumed that a pathogen’s spillover rate is predictive of host jump risk. Here, we develop model that uses a Bayesian framework to quantify the probability that a particular pathogen will host jump within a given future time interval. This model only depends on the rates at which spillover events occur in the past and future, and our uncertainty regarding the probability that a single spillover will result in a host jump. We derive an analytical solution to this model, which shows that when past and future spillover rates are linearly correlated, a pathogen’s inherent rate of spillover is a poor predictor of host jump risk owing to two complementary factors. First, the pathogens that pose the greatest host jump risk may be either those that spillover extremely rarely or extremely commonly, with the result depending on our prior uncertainly in the probability that a spillover will result in a host jump. Second, as spillover rate becomes large, the increase in the number of opportunities for a host jump to occur conferred by the increase in the number of spillover events is exactly balanced by a decrease in the probability of success for each individual spillover event. Therefore, we show that spillover rate is not a useful metric for evaluating host jump risk.
| Expedited Notification | No thanks, I do not require Expedited Notification |
|---|
