https://github.com/theislab/disent
Out-of-distribution prediction with disentangled representations for single-cell RNA sequencing data
https://github.com/theislab/disent
Last synced: 10 months ago
JSON representation
Out-of-distribution prediction with disentangled representations for single-cell RNA sequencing data
- Host: GitHub
- URL: https://github.com/theislab/disent
- Owner: theislab
- License: bsd-3-clause
- Created: 2019-11-07T15:33:05.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2021-05-08T12:21:52.000Z (about 5 years ago)
- Last Synced: 2024-12-30T12:46:52.467Z (over 1 year ago)
- Language: Jupyter Notebook
- Homepage:
- Size: 30.1 MB
- Stars: 6
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
## Out-of-distribution prediction with disentangled representations for single-cell RNA sequencing data
The pdf for the paper can be found [here](https://github.com/theislab/disent/raw/master/paper.pdf).
### Usage
Reproducing results from Jupyter Notebooks:
* JN1_train.ipynb - loads, normalizes and trains the data
* JN2_kang_analysis.ipynb - further analyses the trained model for Kang dataset. It calculates both types of disentanglement scores and also visualizes the latent space as well as gene-feature space plots.
* JN3_dentate_analysis.ipynb - further analyses the trained model for Dentate Gyrus dataset.
* JN4_Out of Distribution Prediction.ipynb - implements OOD Prediction and analyses the results further
* Model Comparisons folder provides the notebooks to reproduce the model architecture comparison plots.
### Misc
* The **environment.yml** file specifies the conda environment in which the project was run.
* Data can be accessed from [here](https://drive.google.com/open?id=1ywXG0K-_nuqnFL8u3x4klsPNOrEdeg27)
## Reference
please consider citing
```
@inproceedings{lotfollahi2020out,
title={Out-of-distribution prediction with disentangled representations for single-cell RNA sequencing data},
author={Lotfollahi, Mohammad and Dony, Leander and Agarwala, Harshita and Theis, Fabian},
booktitle={ICML 2020 Workshop on Computational Biology (WCB) Proceedings Paper},
volume={37},
year={2020}
}