scverse/scvi-tools

Deep probabilistic analysis of single-cell and spatial omics data

What it solves

scvi-tools provides a specialized framework for the probabilistic modeling and analysis of single-cell omics data. It helps researchers perform complex computational tasks such as dimensionality reduction, data integration, automated annotation, and spatial deconvolution on large-scale biological datasets.

How it works

The library is built on top of PyTorch and AnnData, leveraging powerful machine learning frameworks like PyTorch Lightning and Pyro. It offers a high-level API that integrates directly with Scanpy, allowing users to run advanced models with GPU acceleration and standard save/load functionality. For developers, it provides modular building blocks to create and deploy new probabilistic models.

Who it’s for

It is designed for computational biologists and researchers working with single-cell, multi-omics, or spatial omics data who need robust, scalable tools for statistical analysis and model development.

Highlights

  • Supports a wide range of analysis tasks including doublet detection and factor analysis.
  • Built-in GPU acceleration for high-performance computing.
  • Seamless integration with the Scanpy ecosystem.
  • Modular architecture designed for the rapid development of novel probabilistic models.
  • Part of the scverse project, ensuring compatibility with standard single-cell analysis workflows.

Related

  • Project
    TorchSpatiotemporal/tsl

    A PyTorch-based library for neural spatiotemporal forecasting and data processing, focusing on Graph Neural Networks to simplify model prototyping and benchmarking.

  • Project
    pyro-ppl/pyro

    A deep probabilistic programming library built on PyTorch that allows users to create and scale universal probabilistic models for large datasets.

  • Project
    meta-pytorch/botorch

    A PyTorch-based library for Bayesian Optimization that provides a modular framework for composing probabilistic models and acquisition functions.

  • Project
    jacobgil/pytorch-grad-cam

    A PyTorch library for explainable AI in computer vision that provides a wide range of pixel attribution methods to visualize and diagnose model predictions.

  • Project
    sktime/skpro

    A Python library for supervised probabilistic prediction that provides tools for tabular regression, survival analysis, and uncertainty quantification with scikit-learn compatible interfaces.