This repository contains the Python implementation for GraphSCI. Further details about GraphSCI can be found in our paper:
Jiahua Rao, Xiang Zhou, Yutong Lu, Huiying Zhao, Yuedong Yang. Imputing Single-cell RNA-seq data by combining Graph Convolution and Autoencoder Neural Networks
=================
- TensorFlow (1.0 or later)
- python 3.6
- scikit-learn
- scipy
- scanpy
see our manuscript and tutorial for more details.
=================
Preprocess the gene expression matrix and construct the input of gene-to-gene relationships
python train.py --adata ./splatter_data/counts_simulated_dataset3_3000x3000_dropout0.30.h5ad --adj --learning_rate 1e-3 --epochs 100 --hidden1 32 --hidden2 64 --batch_size 50 --dropout 0.1
=================
python train.py
If you want to use our codes and datasets in your research, please cite our paper:
@article{rao2020imputing,
title={Imputing single-cell RNA-seq data by combining graph convolution and autoencoder neural networks},
author={Rao, Jiahua and Zhou, Xiang and Lu, Yutong and Zhao, Huiying and Yang, Yuedong},
journal={biorxiv},
year={2020},
publisher={Cold Spring Harbor Laboratory}
}