🧬 Heterogeneous Biomedical Entity Representation Learning for Gene–Disease Association Prediction

<div align="left">

</div>

---

The FusionGDA model employs a contrastive learning–based training strategy to refine the representations of genes and diseases derived from pre-trained language models, and further introduces an attention-based fusion module to integrate these representations for more accurate gene–disease association (GDA) prediction.

---

🧩 Framework

<div align="center"> <img src="Figure/FusionGDA.jpg" alt="FusionGDA Framework" width="650"/> </div>

---

⚙️ Installation

# Download and install Anaconda
wget https://repo.anaconda.com/archive/Anaconda3-latest-Linux-x86_64.sh
bash Anaconda3-latest-Linux-x86_64.sh -b
rm Anaconda3-latest-Linux-x86_64.sh
export PATH="/root/anaconda3/bin:$PATH"

# Update Anaconda packages
conda update --all

# Install PyTorch with CUDA support (adjust CUDA version if needed)
conda install pytorch torchvision torchaudio cudatoolkit=11.3 -c pytorch

# Install dependencies
pip install wandb PyTDC lightgbm pytorch-metric-learning

---

🚀 Execution

Ensure you are in the directory:

~/dpa_pretrain/scripts

Then adjust parameters as required.

🔹 Pre-training Phase

bash run_pretrain_gda_ml_adapter_infoNCE.sh

🔹 Fine-tuning Phase

TDC Dataset

bash run_finetune_gda_lightgbm_infoNCE_tdc.sh

DisGeNET Dataset

bash run_finetune_gda_lightgbm_infoNCE.sh

Results can be tracked through your Weights & Biases account.

---

📊 Datasets

All datasets are obtained from the following public biomedical repositories:

• TDC: https://tdcommons.ai/
• DisGeNET: https://www.disgenet.org/

The specific versions used in our experiments are stored in the shared Drive:
👉 Shared Drive Link

---

📝 Citation

If you find FusionGDA useful for your research, please cite:

@article{meng2024heterogeneous,
  title={Heterogeneous biomedical entity representation learning for gene-disease association prediction},
  author={Meng, Zhaohan and Liu, Siwei and Liang, Shangsong and Jani, Bhautesh and Meng, Zaiqiao},
  journal={Briefings in Bioinformatics},
  volume={25},
  number={5},
  pages={bbae380},
  year={2024},
  publisher={Oxford University Press}
}

---

<div align="center">

🧠 Developed by the AI4BioMed Lab,
School of Computing Science, University of Glasgow, UK 🇬🇧

</div>