From data to impact - a practical guide to inference for infectious disease dynamics
This course offers a hands-on journey into the intervention quantification for infectious disease dynamics. Following two complementary models approaches, we show how to integrate various data sources in order to build population-level models of disease dynamics. Parameter inference and counterfactual modeling on a historical dataset of malaria cases in Sri Lanka will uncover strengths and limitations of modeling approaches to real-world situations where it is not possible to have a well-designed control.
Instructor: Monica Golumbeanu, Christian Selinger
Term: Spring
Location: Swiss TPH, Seminarraum 4
Time: Wednesdays, 4:15-6:00 PM
Course Overview
This course provides a comprehensive introduction to data science principles and practices. Students will:
-
Integrate multimodal data sources: Work with epidemiological, climate and genomic data to study infectious disease dynamics and evaluate intervention impact.
-
Apply different modeling approaches: Use time series models, mechanistic mathematical models, and parameter inference techniques on real-world infectious disease data.
-
Interpret & communicate results: Analyze model outputs critically and draw meaningful conclusions across different epidemiological contexts and applications.
-
Perform model selection nad parameter inference: Critically evaluate model assumptions, limitations, and appropriateness for specific research questions and contexts.
Prerequisites
- Basic programming knowledge in R
- Introductory statistics
- Comfort with basic algebra
Textbooks
- Modeling Infectious Disease by Matt Keeling and Pejman Rohani
- Epidemics: Models and Data Using R by Ottar Bjørnstad
- Applied Regression Analysis and Generalized Linear Models by John Fox
Further readings
- The rules of contagion by Adam Kucharski
Schedule
| Week | Date | Topic | Materials |
|---|---|---|---|
| 1 | Mar 3 | Introduction & time series analysis Time series analysis of malaria data from Sri Lanka, additive vs multiplicative, auto-correlation, seasonality | |
| 2 | Mar 11 | Time series analysis, continued Time series analysis of malaria data from Sri Lanka, additive vs multiplicative, auto-correlation, seasonality | |
| 3 | Mar 18 | Statistical inference for malaria case count data Poisson regression, generalized linear models, model selection | |
| 4 | Mar 25 | Statistical inference for malaria case count data, continued Model validation, prediction, intervention models, counterfactuals | |
| 5 | Apr 1 | Statistical inference for malaria case count data, introduction to case study Malaria data for Sri Lanka, climate data, intervention data | |
| 6 | Apr 8 | Case study - retrospective impact quantification with Poisson regression models Group work in class | |
| 7 | Apr 15 | Case study - presentations and role play Group work presentations | |
| 8 | Apr 22 | Simulating disease transmission Feedback, comments on causality, analog simulator of epidemic, chain binomial model | |
| 9 | Apr 29 | Disease transmission models Flow diagrams, compartmental models, ordinary differential equations, time-dependent coefficients | |
| 10 | May 6 | Disease transmission models for vector-borne disease & parameter inference Flow diagrams, compartmental models, ordinary differential equations, likelihood | |
| 11 | May 13 | Disease transmission models for Sri Lanka Data integration, differential equations | |
| 12 | May 20 | Case study - retrospective impact quantification with disease transmission models Data integration, differential equations, maximum likelihood, counterfactuals | |
| 13 | May 27 | Casy study - presentations, comparison between modeling approaches, feedback Presentation and discussion |