A powerful and flexible machine learning platform for drug discovery
TorchDrug
TorchDrug is a PyTorch-based machine learning toolbox designed for several purposes.
- Easy implementation of graph operations in a PyTorchic style with GPU support
- Being friendly to practioners with minimal knowledge about drug discovery
- Rapid prototyping of machine learning research
Installation
TorchDrug is compatible with Python >= 3.5 and PyTorch >= 1.4.0.
From Conda
conda install -c milagraph -c conda-forge torchdrug
From Source
TorchDrug depends on rdkit, which is only available via conda.
You can install rdkit with the following line.
conda install -c conda-forge rdkit
git clone https://github.com/DeepGraphLearning/torchdrug
cd torchdrug
pip install -r requirements.txt
python setup.py install
Quick Start
TorchDrug is designed for human and focused on graph structured data.
It enables easy implementation of graph operations in machine learning