# scverse/scvi-tools

Deep probabilistic analysis of single-cell and spatial omics data

Repository: https://github.com/scverse/scvi-tools
Canonical: https://ross.abutalabs.com/products/scvi-tools
Homepage: http://scvi-tools.org/
Language: Python
License: BSD-3-Clause
License Family: permissive
Topics: scrna-seq, variational-bayes, variational-autoencoder, cite-seq, single-cell-genomics, single-cell-rna-seq, deep-generative-model, human-cell-atlas, scverse, deep-learning
Last push: 2026-08-25T14:35:41+00:00

## Health v2 (maintenance only)
Score: 93/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 80, longevity 100
- inputs: {"age_days": 3283, "days_push": 8, "days_rel": 55, "gap_med": 38.5, "n_releases_24m": 19}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1681, forks 471 (observed 2026-08-28T04:05:21.781645+00:00)

## What it is
scvi-tools is a Python library for deep probabilistic modeling and analysis of single-cell and spatial omics data, built on PyTorch, PyTorch Lightning, Pyro, and AnnData. It provides ready-to-use models (scVI, scANVI, totalVI, Stereoscope) for tasks like dimensionality reduction, data integration, and automated annotation, plus building blocks for developing new probabilistic models.

## Use cases
- integrate multiple single-cell RNA-seq datasets
- reduce dimensionality of scRNA-seq data
- automatically annotate cell types
- detect doublets in single-cell data
- deconvolve spatial transcriptomics data
- analyze CITE-seq multi-omic data
- perform factor analysis on single-cell omics
- develop custom variational autoencoder models for genomics

## When to choose
- analyzing single-cell or spatial omics data with deep generative models
- batch-correcting and integrating heterogeneous scRNA-seq datasets
- building novel probabilistic models on top of a shared PyTorch/AnnData infrastructure
- needing GPU-accelerated, scalable inference for large single-cell datasets

## When to avoid
- you need simple exploratory analysis without deep learning overhead
- your data is bulk omics rather than single-cell
- you require a non-Python workflow without any Python integration
- you lack GPU resources and datasets are very large

## Facets
- artifact type: library
- maturity: stable
- function: machine-learning, deep-learning, data-science, llm-training
- domain: bioinformatics, machine-learning, data-science
- platform: python, cross-platform
- tags: single-cell, omics, variational-inference, pytorch, anndata, scanpy, scrna-seq, spatial-omics, generative-models, gpu

## Member repositories
- scverse/scvi-tools (main) score 93

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:21.781645+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:41:26.671710+00:00, confidence not recorded.
  - readme: https://github.com/scverse/scvi-tools (fetched 2026-08-28T04:05:21.781645+00:00, sha 02f8bb7ad6db)
  - homepage: http://scvi-tools.org/ (fetched 2026-08-29T11:15:09.188758+00:00, sha 4b41c9795d70)
  - registry_pypi: https://pypi.org/pypi/scvi-tools/json (fetched 2026-08-29T11:15:09.192708+00:00, sha 56cd0f23f86b)
- Data as of 2026-08-30T08:39:29.467469+00:00.