Benchmarking Graph Learning for Drug-Drug Interaction Prediction

Environment Setup

Make sure you have Anaconda or Miniconda installed on your system before you start. This guide is designed for systems with a CUDA-enabled GPU.

# Create and activate a new environment
conda create -n DDIBench python=3.8

# install dependencies
pip install torch==1.12.0+cu113 torchvision==0.13.0+cu113 torchaudio==0.12.0 --extra-index-url https://download.pytorch.org/whl/cu113
pip install -r requirements.txt

Dataset Preparation

This work uses 2 different datasets, which can be downloaded from this link. Please unzip the downloaded files into a folder naned data within the directory.

Benchmarking

Running the model training step.

python main.py --model MLP --dataset drugbank

• Model choice: CompGCN, SkipGNN, ComplEx, MSTE, MLP, KGDDI, CSMDDI, HINDDI, Decagon, SumGNN, KnowDDI, EmerGNN
• Dataset choice: drugbank, twosides
• Other hypermeters can also be adapted.

Dataset Information

Dataset	#Nodes	#Relations	#Triplets
DrugBank	1710	86	188509
TWOSIDES	645	209	116650