STARNet requires Python 3.11. Other Python versions are not claimed as supported unless separately validated.
We recommend installing STARNet within a conda environment to avoid dependency conflicts.
STARNet is developed and tested on Linux. macOS and Windows (WSL) may work but are not actively validated.
Currently, STARNet is not available on PyPI. Install directly from the GitHub repository:
1. Clone the Repository
git clone https://github.com/DBinary/STARNet.git2. Create and Activate a Conda Environment
conda create -n starnet python=3.11
conda activate starnet3. Install STARNet
cd STARNet
pip install -e .For a validated, reproducible environment with pinned dependencies:
conda env create -f environment-review.yml
conda activate starnet-review
python -m pip install -e .This method uses the environment-review.yml file bundled in the repository and is the preferred workflow for reproducing manuscript results.
After installation, verify that STARNet imports correctly:
import STARNet as STOn some systems, the system libstdc++ may be picked before the active conda environment, causing errors for optional genomics tooling. If this happens, export the active environment library path before running GRN inference:
export LD_LIBRARY_PATH="$CONDA_PREFIX/lib:$LD_LIBRARY_PATH"For GPU acceleration, install the appropriate CuPy build for your CUDA toolkit after installing STARNet.