A pytorch implementation of trVAE (transfer Variational Autoencoder). trVAE is a deep generative model which learns mapping between multiple different styles (conditions). trVAE can be used for style transfer on images, predicting single-cell perturbations responses and batch removal.
To install the latest version from PyPI, simply use the following bash script:
pip install trvaepor install the development version via pip:
pip install git+https://github.com/theislab/trvaep.gitor you can first install flit and clone this repository:
pip install flit
git clone https://github.com/theislab/trvaep
cd trvaep
flit install-
For simple perturbation prediction and batch-removal check this example with interferon (IFN)-β stimulation from Kang et al..
-
For multi condition perturbation prediction and batch-removal check this example with multiple infections from Haber et al..
In order to reproduce paper results visit here.