Medical School Office Building (MSOB)
|DATE:||October 19, 2017|
|TIME:||1:30 - 2:50 pm|
|TITLE:||Exploring multiple approaches to inferring environmental drivers of dengue dynamics|
Department of Biology, Stanford
Infectious disease dynamics are highly nonlinear, and their responses to the environment involve complex, nonlinear interactions between infectious agents (bacteria, virus, parasites), hosts (people, plants, animals), and the environment (temperature, habitat). It is critical to understand environmental drivers of infectious disease in order to predict, detect, and prevent epidemics, particularly for diseases with no effective vaccines or treatments. In this talk, I will discuss approaches we have taken to infer the environmental drivers of dengue dynamics, a mosquito-borne viral disease that is estimated to cause 390 million cases per year. We first built mechanistic (bottom-up) models of temperature impacts on dengue transmission by considering the impact of temperature on the mosquito and virus traits that drive transmission. Second, we used epidemiological forecasting model approaches that include both mechanistic models with environmental covariates (i.e., generalized linear models) and phenomenological machine learning models (i.e., Gaussian process models). Finally, I introduce a new approach we are taking to infer the causal relationship between weather and dengue dynamics using convergent cross-mapping: an approach pioneered in infectious disease dynamics by George Sugihara and colleagues. I am particularly interested in getting feedback on these complementary approaches and how to best move forward with statistical (and particularly, causal) inference in complex, nonlinear systems.
LR Johnson, RB Gramacy, J Cohen, E Mordecai, C Murdock, J Rohr, SJ Ryan, AM Stewart-Ibarra and D Weikel. Phenomenological forecasting of disease incidence using heteroskedastic Gaussian processes: a dengue case study. arXiv:1702.00261v3 [stat.AP] Aug 2017.
ER Deyle, MC Maher, RD Hernandez, S Basu, and G Sugihara. Global environmental drivers of influenza. PNAS vol. 113 no. 46, 13081-13086, doi:10.1073/pnas 1607747113.