Dynamic Inference

Professor Pejman Rohani (University of Georgia) is the Lead on Research Project 2.

The Investigators are based at University of Georgia and University of Michigan. 

The main objective is the development of cutting-edge statistical inference methods and their application to high-dimensional disease incidence data.

The objectives

  1. To establish epidemiological, immunological, and climactic determinants of dengue transmission. This will involve confronting mathematical models that encapsulate specific competing hypotheses with dengue incidence data from Thailand and Mexico, using methods described under objective 3. This will establish (a) duration and strength of homologous immunity, (b) the transmission consequences of antibody dependent enhancement, (c) evidence for climatic drivers of seasonality in dengue transmission, and (d) the determinants of spatial diffusion in dengue. This sub-project will lead to the integration of data across multiple scales, from individual-level immunological interactions to climate-driven vector dynamics, and will permit the development of predictive, validated models.
  2. To establish epidemiological, immunological, and vaccine-specific determinants of polio transmission. This will involve interrogating historical reports and genetic sequence data from the US  (1920s-1950s) together with contemporary surveillance reports and genetic sequence data from India to examine (a) the duration and nature of infection- and vaccine-derived immunity, (b) age-specific and seasonal drivers of transmission and (c) safe strategies for endgame management.  
  3. To develop, implement, and disseminate novel statistical inference methods for high-dimensional systems. To accomplish objectives 1 and 2, the investigators will continue the development of powerful computational methods for inference from dynamical data, to extend, employ, and disseminate these methods to high-dimensional problems posed by data stratification by age, space, genetics and strain.