GEMDeCan: Gene Expression and Methylation based Deconvolution for Cancer
You will find here a snakemake pipeline to perform immune cell type deconvolution using 6 methods and 12 signatures based on gene expression and methylation data.
- inputs : Directory where your inputs are located along with the signatures files used in the pipeline.
- results : Directory where your output will be saved.
cibersortx/is where CibersortX results are saved,deconvolutions/is where results from all the other methods without CibersortX are saved andall_deconvolutions/is where all deconvolution results merged are saved. - scripts : Directory where you will find all the scripts produce to run the pipeline
- Gene expression matrix: TPM normalized in .txt format
Format specification
gene Sample1 Sample2 Sample3
g1 1.8 5.2 4.1
g2 5.7 8.3 2.0
g3 6.2 3.1 9.2
g4 7.2 9.1 0.6
Once this format is achieved, you can save the input file from R using:
write.table(counts.TPM.normalized, "GEMDeCan_deconvolution/inputs/file_name.txt", quote = F, sep = "\t", row.names = F)
Note that the input file should be saved in inputs/ directory. Additionally, go to the snakemake.yml file and add the name of your input file (without extension) in the Mixtures section.
Snakemake allows for a very efficient and user friendly way of using pipelines. It is designed so all you need to install is mamba which is required to install Snakemake Note that officially only Linux is supported for this pipeline. This also requires an Internet connection in order to use mamba auto-generated environments for all necessary software and packages.
Mamba
- Linux
curl -L -O "https://github.com/conda-forge/miniforge/releases/latest/download/Mambaforge-$(uname)-$(uname -m).sh"bash Mambaforge-$(uname)-$(uname -m).sh
Snakemake
- Linux
conda activate basemamba install snakemake -n base -c conda-forge -c bioconda
Linux packages
sudo apt install build-essential
sudo apt-get install r-base-dev
sudo apt-get install -y libtiff5-dev zlib1g-dev libpng-dev libxml2-dev libharfbuzz-dev libfribidi-dev libcurl4-openssl-dev libfontconfig1-dev liblapack-dev libopenblas-dev
Docker (If not already installed)
- Linux
sudo apt install ca-certificates curl gnupg lsb-releasesudo mkdir -p /etc/apt/keyringscurl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo gpg --dearmor -o /etc/apt/keyrings/docker.gpgsudo chmod a+r /etc/apt/keyrings/docker.gpgecho "deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.gpg] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable" | sudo tee /etc/apt/sources.list.d/docker.list > /dev/nullsudo apt updatesudo apt install docker-ce docker-ce-cli containerd.io
R packages
After running GEMDeCan succesfully for the first time, go to the files scripts/run_cibersort_local_container.R, scripts/merge_all_deconvolutions.R and scripts/deconvolution/deconvolution_algorithms.R and uncomment the first lines to avoid re-looking for them in the future.
- Input file should be save in folder
inputs/ - Go to the file
snakemake.ymland add the name of your input file in the sectionMixtures
Either sudo snakemake --cores 8 all or sudo path_where_snakemake_is_located --cores 8 all depending if snakemake has been installed locally or not.
CIBERSORTx is included in the deconvolution methods in the GEMDeCan pipeline, but it's not run by default as it's not an open-source program.
To run it, please ask for a token in CibersortX and create a file named credentials.txt with your username and password on separate lines.
MAIL: [email protected]
TOKEN: token_passowrd123
To cite the pipeline, refer to our paper on BiorXiv :
GEM-DeCan: Improved tumor immune microenvironment profiling through novel gene expression and DNA methylation signatures predicts immunotherapy response
Ting Xie, Jacobo Solorzano, Miguel Madrid-Mencía, Abdelmounim Essabbar, Julien Pernet, Mei-Shiue Kuo, Alexis Hucteau, Alexis Coullomb, Nina Verstraete, Olivier Delfour, Francisco Cruzalegui, Vera Pancaldi
bioRxiv 2021.04.09.439207; doi: https://doi.org/10.1101/2021.04.09.439207