This is an official implementation of paper "GeSubNet: Gene Interaction Inference for Disease Subtype Network Generation". https://arxiv.org/abs/2410.13178
Our work was selected as Oral Presentation at ICLR 2025!
GeSubNet is built based on PyTorch. You can install PyTorch following the instructions in PyTorch. For example:
pip install torch==1.10.0+cu111 torchvision==0.11.0+cu111 torchaudio==0.10.0 -f https://download.pytorch.org/whl/torch_stable.htmlAfter ensuring that PyTorch is installed correctly, you can install other dependencies via:
pip install -r requirements.txtWe have prepared datasets for four cancers: BRCA, GBM, OV, and LGG. The dataset is in the in the directory ./data/.
For GeSubNet:
The main code is ./GeSubNet_main.py.
The available models for Patient-M and Graph-M are in the directory ./model/.
The utils needed for training are ./utils.py.
Log files will be generated and updated in ./log/ during the training process.
More results of the BRCA case analysis are in the directory ./Case Analysis/.
If you find our work useful in your research, please consider citing:
@inproceedings{
yang2025gesubnet,
title={GeSubNet: Gene Interaction Inference for Disease Subtype Network Generation},
author={Ziwei Yang and Zheng Chen and Xin Liu and Rikuto Kotoge and Peng Chen and Yasuko Matsubara and Yasushi Sakurai and Jimeng Sun},
booktitle={The Thirteenth International Conference on Learning Representations},
year={2025}
}