conda create -y -n ONTraC python=3.11 # ONTraC supports Python 3.10, 3.11, and 3.12 for now
conda activate ONTraC
pip install ipykernel
python -m ipykernel install --user --name ONTraC --display-name "Python 3.11 (ONTraC)"
pip install ONTraC
# Use this command if you want to visualise the results by `ONTraC_analysis`.
pip install "ONTraC[analysis]"
NOTE: For ARM-based macOS, we recommend installing via pip for now, as the dependency package, pytorh-geometric, does not have a conda build for it. NOTE: For x86-based macOS, conda install only support Python 3.10 and 3.11 for now. WARNING: Installing with conda can be very slow.
conda install -c gyuanlab -c pytorch -c pyg -c default -c nvidia -c conda-forge ontrac
git clone [email protected]:gyuanlab/ONTraC.git .
cd ONTraC
pip install .
# Use this command if you want to visualise the results by `ONTraC_analysis`.
pip install ".[analysis]"