All courses take the form of tutorials. Statistical methods and concepts are motivated by illustrative situations in a large scope of applications in life sciences. Basic functions in R are introduced to implement those methods.
First steps in data analysis with R
Short description: The course introduces the basics of inferential statistics. It essentially provides tools to answer the question “Is there an effect of this on that?” using data. The scope of statistical methods in the course is limited to common variants of the former question: “Are there differences across group means of that?” and “Is there a linear relationship between this and that?”.
Keywords: Hypothesis testing, Estimation, One-way Analysis of variance, F-test, t-test, Simple linear regression model
Statistics for decision making
Short description: In many situations, especially in biological applications, data analysis aims at understanding why some profiles have a higher probability than others to belong to a class of interest. Typically, which clinical profiles are more likely to be affected by a disease? The course introduces statistical models to answer such questions and the following sub-questions: Does a covariate significantly explain variations between class probabilities? How to select a minimal and sufficient subset of covariates explaining the variations between class probabilities? How to use a profile of covariates to predict the membership to a given class?
Keywords: Logistic linear regression model, Analysis of deviance testing, AIC, BIC, Classification errors.
Short description: In molecular biology, genomics aims at understanding how the genome works using whole-genome measurements obtained by high-throughput biotechnology. Analyzing those massive data raises many specific statistical issues. The most common of those issues are addressed in this introductory course.
High-dimensional regression and classification modeling for biological data analysis
Short description: High-throughput technology (imaging, spectroscopy, etc.) is now currently used for many biological applications. Using high-dimensional data generated by those technology to make prediction rules require ad-hoc methods. The most commonly used of these methods are introduced in the course.
Introduction to geostatistics.
Short description: The course introduces the prediction issue with geolocalized data (kriging) using a spatial stationary dependence model (variogram modeling). Slides, R script, Data and Handbook.
Introduction to biomathematics.
Short description: The purpose of the course is to introduce basic mathematical concepts motivated by biological modeling applications. Functional analysis slides, Enzyme kinetics.