RNA-protein interactions are critical to various life processes, including fundamental translation and gene regulation. Identifying these interactions is vital for understanding the mechanisms underlying life processes. Then, ZHMolGraph is an advanced pipeline that integrates graph neural network sampling strategy and unsupervised large language models to enhance binding predictions for novel RNAs and proteins.

• We provide the script and model for validating the results of ZHMolGraph. Any machines with a GPU and an Ubuntu system should work.

• We recommend using Anaconda to create a virtual environment for this project.

• you will need a major software package: pytorch. The following commands will create a virtual environment and install the necessary packages. Note that we install the GPU version of PyTorch (torch==1.8.1+cu11) for training purpose.

conda create -n ZHMolRPGraphPytorch-1.8 python=3.8
conda activate ZHMolRPGraphPytorch-1.8
pip install tqdm
pip install torch==1.8.1+cu111 torchvision==0.9.1+cu111 torchaudio==0.8.1 -f https://download.pytorch.org/whl/torch_stable.html
pip install tensorboard
pip install jupyter

• All Python modules and corresponding versions required for ZHMolGraph are listed here: requirements.txt

• Use pip install -r requirements.txt to install the related packages.

All data files are available here: https://zenodo.org/record/

• data/interactions/: Contains RNA-protein interactions data.
• data/Mol2Vec/: Contains the embeddings from LLMs.
• trained_model/: Contains the trained model.

Here we describe the Jupyter Notebooks scripts used in ZHMolGraph.

1_NPInter2_result_validation.ipynb: We execute a five-fold cross-validation on benchmark dataset NPInter2.

2_RPI7317_result_validation.ipynb: We execute a five-fold cross-validation on benchmark dataset RPI7317.

3_TheNovel_unknown_dataset_validation.ipynb: We conduct testing on the unseen nodes dataset NPInter5 using the trained models generated from the five-fold cross-validation on the benchmark dataset NPInter2.

If you have any comments, questions or suggestions about the ZHMolRPGraph, please contact:

• Yunjie Zhao E-mail: [email protected]
• Haoquan Liu E-mail: [email protected]