This repository contains the code for the research paper titled " Personalized decision support system for tailoring IgA nephropathy treatment strategies ". The project is divided into three main components: Autoencoder, Traditional Graph Feature Engineering, and Network Biomarker Construction.
This project aims to optimize treatment strategies for IgAN patients using network biomarkers. We employ a combination of autoencoders, traditional graph feature engineering techniques, and a novel network biomarker construction method to identify key biomarkers and enhance decision-making.
-
Clone the repository:
git clone https://github.com/your-username/renal-replacement-therapy.git -
Navigate to the project directory:
cd renal-replacement-therapy -
Install the required dependencies:
pip install -r requirements.txt
-
Run the autoencoder model script:
python autoencoder/model.ipynb -
Execute the graph feature extraction script:
python graph_features/feature_extraction.ipynb -
Construct network biomarkers using:
python network_biomarkers/biomarker_construction.ipynb
The autoencoder/ directory contains the script for building and running the autoencoder model used for dimensionality reduction and feature extraction.
model.ipynb: Defines and trains the autoencoder model.
The graph_features/ directory contains the script for extracting traditional graph features from the data.
feature_extraction.ipynb: Extracts and analyzes graph-based features.
The network_biomarkers/ directory includes the script for constructing and analyzing network biomarkers.
biomarker_construction.ipynb: Constructs network biomarkers from the extracted features.
matplotlib==3.7.1
networkx==3.3
numpy==1.25.2
pandas==2.0.3
scipy==1.11.4
torch==1.9.0+cu102
Contributions are welcome! Please open an issue or submit a pull request for any improvements or additions.
This project is licensed under the MIT License