West Nile

As part of the ARS Predictive Disease Ecology Grand Challenge Project led by Drs. Deb Peters and Luis Rodriguez, a spatiotemporal disease model was developed to forecast future West Nile Disease (WND) outbreaks in horses across the continental US. Postdoc Dr. John Humphreys led the analysis to predict the distribution and timing of future WND outbreaks. The Centers for Disease Control and Prevention (CDC) records provided the count of veterinary-reported WND cases for horses between 2000 – 2018 aggregated by county. The team used the USDA National Agricultural Statistics Service database to map horse populations, incorporated CDC mosquito surveillance reports to identify insect vector ranges, and analyzed more than 10 million bird occurrence records from the Cornell Laboratory of Ornithology to map the distributions of avian species known to host WNV. These datasets allowed the model to link the at-risk livestock population (horses) to times and locations with both the WNV reservoirs (birds) and the WNV vectors (mosquitos) that transmit the virus between those reservoirs and livestock. The research team applied a Bayesian hierarchical modeling framework to construct the model and specified that the prediction for any one location be dependent on the disease risk estimated for surrounding areas and past times (manuscript is in preparation).

​

To read more, see the SCINet Newsletter story covering the project here.