Multivariate analysis for -omics data integration: principles and applications

Date 29 November - 2 December 2021
Speaker(s) Ricard Argelaguet, Sébastien Déjean, Pietro Ferrari, Vivian Viallon, Mathieu Foll, Reza Salek, Florence Mehl, Thuong Van Du Tran
Cancellation deadline 2 Nov 2021
City Lyon

This course will take place in Lyon with rules complying to COVID situation (distances and hygiene), but in the event of restrictions preventing the event to take place in person, the participants will be informed about a probable streamed version.

IARC (International Agency for Research on Cancer), the doctoral program Staromics of the CUSO (Conférence Universitaire de Suisse Occidentale) and SIB are pleased to co-organize this course. Part of the places will be kept in priority for members of IARC and Staromics.


Researchers often have access or generate multiple -omics data within a single study. Although each –omics data has been traditionally analysed in isolation, combining possibly complementary data can yield a better understanding of the mechanisms involved in the biological processes. Several integrative approaches are now available to combine such data, which can be regarded as extensions of the standard Principal Component Analysis (PCA).

In this 3.5 day workshop, we will provide an overview of omics data structures, and present different statistical approaches, from simple PCA to more advanced multi-omics methods. For each method, we will cover both its principle and practical aspects. In particular, the course will include dedicated sessions covering the application of the MOFA and MixOmics packages by their respective developers. We plan to have an interactive course with substantial time allocated for discussions.


Day 1: Overview of omics data structures and standard dimension reduction techniques.

Day 2: MOFA (Multi-Omics Factor Analysis) - unsupervised multiomics data integration.

Day 3: MixOmics package - supervised multiomics data integration methods.

Day 4 (1/2 day): Case studies.