scvi-tools

Name: scvi-tools
Availability: InStock
Author: rubensliv

Model and analyze single-cell data with probabilistic VAEs for batch-corrected latent representations.

15|2|Updated Dec 17, 2025

Search Tags:#multimodal #single-cell #batch-correction #scvi-tools #differential-expression #variational-inference

Ready to run

npx skills add https://github.com/rubensliv/k-dense-ai --skill scvi-tools-rubensliv

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: scvi-tools
Source: https://github.com/rubensliv/k-dense-ai/tree/main/scientific-skills/scvi-tools
Command: npx skills add https://github.com/rubensliv/k-dense-ai --skill scvi-tools-rubensliv

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Bayesian, variational-inference-based models for analyzing and integrating large-scale single-cell data, providing batch-corrected latent representations and probabilistic differential expression.

Core Features & Use Cases

Probabilistic modeling for scRNA-seq, scATAC-seq, and multimodal data (RNA, protein, and chromatin)
Learning batch-corrected latent representations with scVI family models (scVI, SCANVI, TOTALVI, MultiVI, MethylVI, etc.)
Differential expression and differential accessibility analyses within a Bayesian framework
End-to-end workflows including data registration, clustering, trajectory inference, and cross-dataset transfer
Model persistence, GPU acceleration, and integration with Scanpy/AnnData

Quick Start

Train the scvi-tools model on your AnnData to obtain latent representations and perform downstream analyses such as clustering and differential expression.

Frequently Asked Questions about scvi-tools

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I integrate and perform batch correction on scRNA-seq datasets?▼

To perform batch correction on scRNA-seq datasets, this tool trains probabilistic VAEs on raw count inputs to learn batch-corrected latent representations. It supports multiple models like scVI and SCANVI for large-scale data integration and cross-study transfer.

Can I analyze multimodal single-cell data combining RNA and chromatin accessibility?▼

Yes, you can analyze multimodal single-cell data combining RNA and chromatin accessibility. The MultiVI and TOTALVI models process scRNA-seq, scATAC-seq, and multimodal inputs to generate joint latent representations for integrated clustering and annotation.

What is the best way to perform differential expression analysis within a Bayesian framework?▼

The best way to perform Bayesian differential expression analysis is using variational inference models like scVI. This approach calculates probabilistic differential expression directly from learned latent representations, providing statistically grounded results for single-cell data.

Does scvi-tools work with AnnData and support GPU acceleration?▼

Yes, scvi-tools integrates directly with Scanpy and AnnData for data registration and processing. It supports GPU-accelerated training to efficiently model large-scale single-cell datasets and persist trained models for downstream use.

When should I use probabilistic variational inference for single-cell clustering?▼

You should use probabilistic variational inference for single-cell clustering when you need batch-corrected latent representations from raw counts. This approach handles scRNA-seq, scATAC-seq, and multimodal datasets, offering uncertainty quantification for clustering and trajectory inference.