Implementation of AEMDA for inferring potential disease-miRNA associations. Our method contains three sub-models: a disease model for learning representation of diseases, a miRNA model for learning representation of miRNA, and an autoencoder based model for predicting.
- PyTorch 1.0 or higher
- GPU (default).
- download code and data
- execute
python main.py
to train a predictor
Note: If you wanna infer disease-specific miRNAs, you should use the concatenated vector
Please cite our paper if you use this code in your own work:
@article{Ji2021,
author = {Ji, Cunmei and Gao, Zhen and Ma, Xu and Wu, Qingwen and Ni, Jiancheng and Zheng, Chunhou},
doi = {10.1093/bioinformatics/btaa670},
issn = {1367-4803},
journal = {Bioinformatics},
mendeley-groups = {MDA},
month = {jan},
number = {1},
pages = {66--72},
title = {{AEMDA: inferring miRNA–disease associations based on deep autoencoder}},
url = {https://doi.org/10.1093/bioinformatics/btaa670},
volume = {37},
year = {2021}
}