mediation_models

A set of scripts to illustrate mediation models

Context

http://epimed.univ-grenoble-alpes.fr/downloads/dmzfch/presa.pdf

How to start?

Conda environment and working directory

Set up your conda environment as follow:

# conda create -n mediationmodels_env
conda activate mediationmodels_env
# mamba install -c anaconda -c bioconda -c conda-forge -c r  -c brown-data-science r-base r-rmarkdown r-mediation snakemake=7.32.4 python=3.9 graphviz python-kaleido tenacity plotly 
# pip install smgantt==0.0.5

Set up your working directory

# Retrieve script and material from github
git clone https://github.com/fchuffar/mediation_models.git
# Go to your workdir
cd mediation_models

1st simulation

snakemake -s 01st_simu.py --cores 16 -rp clean
snakemake -s 01st_simu.py --cores 16 -rpn
# snakemake -s 01st_simu.py --jobs 50 --cluster "oarsub --project epimed -l nodes=1/core=1,walltime=00:10:00"  --latency-wait 60 -pn

Exercice

Targeted dataset is datasets/sim_res_2000_1_100_1_500_1_50_1_rep1.txt

Launch mediation_analysis_tp.Rmd (rmarkdown::render("mediation_analysis_tp.Rmd"))
Comment figures, models used, nature of the variables CTCF, c3D and mRNA

Study and comment the two other datsets datasets/sim_res_2000_1_100_1_0_1_50_1_rep0.txt and datasets/sim_res_2000_1_100_0_500_1_50_1_rep0.txt

Adapt the mediation_analysis_tp.Rmd script and propose a mediation_analysis_glm.Rmd script better able to handle this type of data.
What is the impact of the changes on the effects of the various mediation scheme models and on the ADE, as well as on the ACME?

CR noté analyse de médiation

Le sujet est : en se fondant sur le travail réalisé et en reprenant le cadre formel d'une analyse statistique vu en M1 :

analysez le nouveau jeu de données datasets/sim_res2.txt ;
concluez, dans cette nouvelle etude, sur l'impact de CTCF sur la transcription médiée par l'organisation 3D du génome ?

Les données sont ici https://github.com/fchuffar/mediation_models/