Application of graphical models to metabolomics

Ce tutorial faisait part de la conference useR2020, et à cause du covid-19 le tutorial a été offert à R-Ladies Strasbourg. Il sera fait en Anglais car les intervenants présenteront sur zoom depuis les États-Unis.

For this tutorial we will give the priority to R-Ladies Strasbourg and the R-community in Strasbourg and surrounding. Thank you for your understanding.

This tutorial will identify a suite of tools for using network models in applied data analyses. Examples will utilize metabolomics data
from ongoing large-scale health research studies, although techniques are transferrable across multiple domains. This workshop will
link network visualizations with statistical analyses of complex metabolomics data and will emphasize detection of network
subcomponents that link to health outcomes. Utility of these tools for ‘story telling’ in complex data settings will be illustrated.
Learning objectives for the workshop are to:

1. Explain the use of network models in metabolomics and other omics studies
2. Apply R programming to analyze metabolomics and other omics network data
3. Visualize network models using R

The tutorial will include didactic instruction, demonstration of code and hands on work time for provided exercises. All exercises will be constructed in an R markdown format. Example code will be extracted from the .Rmd file for easy copy-and-paste execution during the course.

Didactic introduction to network models and metabolomics (30 min)
Demonstration of code for creating basic network data objects in R (20 min)
Hands on work time for first set of exercises (40 min)
Break (30 min)
Didactic introduction to graphical lasso and differential network analysis (30 min)
Demonstration of code for performing graphical lasso and differential network analysis in R (20 min)
Hands on work time for second set of exercises (40 min)

The course will be co-taught by Raji Balasubramanian, PhD, Associate Professor of Biostatistics at University of Massachusetts at Amherst in Amherst, MA and Denise Scholtens, PhD, Professor of Biostatistics at Northwestern University Feinberg School of Medicine in Chicago, IL.
Drs. Balasubramanian and Scholtens are long time R users (15+ years) and statistical collaborators on the development of network methods for metabolomics data analysis. They previously co-taught a similar workshop at Metabolomics 2019 in The Hague, Netherlands. They are both regular faculty instructors in the graduate biostatistics degree programs at their respective institutions. Links to their faculty pages are below.
https://www.umass.edu/sphhs/person/faculty/raji-balasubramanian
https://www.feinberg.northwestern.edu/faculty-profiles/az/profile.html?xid=16163