Installing Other R packages on O2
Please make sure to setup an R Personal Library on O2 to install packages locally.
Here you can find valuable information on how to install R packages requiring some external dependencies.
R package - curl
# Load modules
module purge
module load gcc/9.2.0 R/4.4.0
# Install cURL locally
mkdir $HOME/CURL_LOCAL && cd $HOME/CURL_LOCAL
wget https://curl.se/download/curl-8.7.1.tar.gz
tar xvzf curl-8.7.1.tar.gz && cd curl-8.7.1
./configure --with-openssl --prefix $HOME/CURL_LOCAL
make
make install
# Create a Makevars file with the following output
mkdir ~/.R
echo "CFLAGS=-I/home/$USER/CURL_LOCAL/include" >> ~/.R/Makevars
echo "CXXFLAGS=-I/home/$USER/CURL_LOCAL/include" >> ~/.R/Makevars
echo "LDFLAGS=-L/home/$USER/CURL_LOCAL/lib" >> ~/.R/Makevars
# Launch R and install rtracklayer
R
install.packages('remotes')
remotes::install_github('jeroen/curl@v4.3.2')
The curl
package is needed to install popular R packages such as rtracklayer
and Seurat
(v5). You will need to install an older version of curl until the OS for O2 is updated from CentOS to RedHat.
R package - igraph
# Load Modules
module purge
module load gcc/9.2.0 R/4.3.1 glpk/5.0
# Create ~/.R/Makevars
mkdir ~/.R
echo "CFLAGS+=-I/n/app/glpk/5.0-gcc-9.2.0/include" >> ~/.R/Makevars
echo "CPPFLAGS+=-I/n/app/glpk/5.0-gcc-9.2.0/include" >> ~/.R/Makevars
echo "LDFLAGS+=-L/n/app/glpk/5.0-gcc-9.2.0/lib" >> ~/.R/Makevars
echo "CPATH+=/n/app/glpk/5.0-gcc-9.2.0/include" >> ~/.R/Makevars
# Start R
R
# Install package
install.packages('igraph')
R package - nloptr
You must use an R version compiled against GCC v9.2.0 and load the CMake module. For example:
# Load modules
module purge
module load gcc/9.2.0 R/4.2.1 cmake/3.22.2
# Start R
R
# Install package
install.package('nloptr')
R package - Devtools
Load the git2 module if using an older R version (e.g., 4.0.1):
If using the R>=4.2.1 module, then it is not required to load the git2 module.
R package - Seurat
R packages - SF and Monocle3
R package - ACTIONet
The R package has not been maintained for the last few years; currently it is only possible to use through a singularity container.
R package - InferCNV
R package - Rmpi
The package must be installed using an sbatch job. For example:
1. Create an sbatch script with the following content
2. Submit script via the sbatch command