Teaching
MSc Public Health Data Science – ISPED, University of Bordeaux
Genomics Data : Generation, Management, and Analysis (2024-)
Description: In this 6 hour long course, I give an introduction to genomics data, focusing on the generation and analysis of transcriptomic data. In particular, I discuss
- Quality control, filtration, normalisation, and batch effects
- Exploratory analyses : principal component analysis, hierarchical clustering algorithms
- Differential gene expression analysis
Download (PDF) | Open in new tab
Practical work (questions)Download (.Rmd) | Open in new tab (.html)
Practical work (solutions)Download (.Rmd) | Open in new tab (.html)
Genomics Data : Supervised Learning for High-Dimensional Data (2024-)
Description: In this 6 hour long course, I give an overview on supervised learning in the high-dimensional setting, focusing on regression tasks. I mainly discuss
- Cross-validation for model evaluation and hyperparameter tuning
- Regularised methods (ridge, lasso, and elastic net regression)
- Latent variable approaches (prinipal component regression, (sparse) partial least squares)
- Model interpretability
Download (PDF) | Open in new tab
Practical work (questions)Download (.Rmd) | Open in new tab (.html)
Practical work (solutions)Download (.Rmd) | Open in new tab (.html)
Genomics Data : Test Multiplicity (2024-)
Description: In this 3 hour long course, I provide an overview of the multiple testing problem in high-dimensional settings. In particular, I discuss
- An introduction to frequentist hypothesis testing
- Family-wise error rate and its control (Bonferroni, Holm)
- False discovery rate and its control (Benjamini-Hochberg, Benjamini-Yekutieli)
Download (PDF) | Open in new tab
Practical work (questions)Download (.Rmd) | Open in new tab (.html)
Practical work (solutions)Download (.Rmd) | Open in new tab (.html)
Core Principles of Epidemiology (2024-)
Description: In this 6 hour long course, I give an introduction to epidemiology, discussing
- Measures of association
- Study design
- Diagnostics (sensitivity, specificity, PPV, NPV)
- Causality, types of bias, and directed acylic graphs