Env for reproducing

If you want repeat our whole benchmark and evaluation workflow, please configure environment as follow.

Python env

You can build the python env for snakemake workflow by conda or mamba. We recommend to use mamba to speed up the installation process.

• snakemake

mamba create -p ./conda python==3.8 -y && conda activate ./conda
mamba install -c bioconda -c conda-forge snakemake==7.12.0 tabulate==0.8.10 pandoc -y

git clone git@github.com:gao-lab/SLAT.git
cd SLAT
pip install -e ".[dev,docs]"
pip install pyg_lib torch_scatter torch_sparse torch_cluster torch_spline_conv -f https://data.pyg.org/whl/torch-2.0.0+cu117.html

R env

STEP 1: Install R

You also should configure R environment. We strong recommend to compile a new R-4.1.3 rather than install R in conda.

Warning Please make sure you have deactivated any conda env before using R

cd SLAT/resource
wget https://cran.r-project.org/src/base/R-4/R-4.1.3.tar.gz
tar -xzvf R-4.1.3.tar.gz && cd R-4.1.3 && 
  ./configure --without-x --with-cairo --with-libpng --with-libtiff --with-jpeglib --enable-R-shlib --prefix={YOUR_PATH} &&
   make && make install

STEP 2: Register R kernel

Then register the jupyter kernel for R so snakemake can call R in benchmark workflow.

install.packages('IRkernel')
IRkernel::installspec(name = 'slat_r', displayname = 'slat_r')

STEP 3: Install R packages

At last, please install all R packages we used from renv.lock (see renv).

install.packages('renv')
install.packages('IRkernel') # install IRkernel again inside renv env
renv::restore()

System env

Singularity

You also need install singularity, because we use container to ensure the repeatability of benchmark results.