scVI

Single-cell Variational Inference

• Free software: MIT license
• Documentation: https://scvi.readthedocs.io.

Quick Start

1. Install Python 3.7. We typically use the Miniconda Python distribution and Linux.

2. Install PyTorch. If you have an Nvidia GPU, be sure to install a version of PyTorch that supports it -- scVI runs much faster with a discrete GPU.

3. Install scVI through conda:

   conda install scvi -c bioconda -c conda-forge

   Alternatively, you may try pip (pip install scvi), or you may clone this repository and run python setup.py install.

4. If you wish to use multiple GPUs for hyperparameter tuning, install MongoDb.

5. Follow along with our Jupyter notebooks to quickly get familiar with scVI!

   1. Getting started:

      • data loading
      • basic usage

   2. Analyzing several datasets:

      • harmonization
      • annotation

   3. Advanced topics:

      • interaction with scanpy
      • linear decoder for gene interpretation
      • reproducing results from the scVI paper
      • imputation of unobserved gene expression (gimVI)
      • hyperparameter tuning for scVI with our autotune module

References

Romain Lopez, Jeffrey Regier, Michael Cole, Michael I. Jordan, Nir Yosef. "Deep generative modeling for single-cell transcriptomics." Nature Methods, 2018. [pdf]

Chenling Xu∗, Romain Lopez∗, Edouard Mehlman∗, Jeffrey Regier, Michael I. Jordan, Nir Yosef. "Harmonization and Annotation of Single-cell Transcriptomics data with Deep Generative Models." Submitted, 2019. [pdf]

Romain Lopez∗, Achille Nazaret∗, Maxime Langevin*, Jules Samaran*, Jeffrey Regier*, Michael I. Jordan, Nir Yosef. "A joint model of unpaired data from scRNA-seq and spatial transcriptomics for imputing missing gene expression measurements." ICML Workshop on Computational Biology, 2019. [pdf]