skills.rest
silverstein
silversteinCOMMUNITY
Explore all skills by this author →
Website𝕏@matsilversteinGitHub

scvi-tools

Train scvi-tools models on AnnData to generate batch-corrected latent representations.

21|2|Updated Dec 8, 2025
Search Tags:#single-cell#batch correction#spatial transcriptomics#differential expression#variational inference#multimodal integration
One-click install
npx skills add https://github.com/silverstein/claude-scientific-skills-desktop --skill scvi-tools-silverstein
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: scvi-tools
Source: https://github.com/silverstein/claude-scientific-skills-desktop/tree/main/corpus/scvi-tools
Command: npx skills add https://github.com/silverstein/claude-scientific-skills-desktop --skill scvi-tools-silverstein

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

scvi-tools helps users analyze single-cell omics by providing probabilistic modeling that supports batch correction, dimensionality reduction, and differential expression with principled uncertainty handling.

Core Features & Use Cases

  • Single-cell representation learning: Use variational models to compute latent embeddings for clustering and downstream analysis, e.g., scVI-style workflows.
  • Batch correction & integration: Model technical variation and learn batch-invariant representations for multi-study datasets and covariates like donor, condition, and percent_mito.
  • Model-specific analyses by modality: Apply the right scvi-tools model for RNA-seq, ATAC-seq, multimodal CITE-seq (totalVI), multi-omic integration (MultiVI), spatial transcriptomics (DestVI), and specialized modalities like methylation or cytometry.
  • Probabilistic differential expression/accessibility: Perform differential expression or accessibility using Bayesian posterior sampling and configurable hypothesis testing modes (e.g., “change” with delta).

Quick Start

Use the scvi-tools skill to train a model on raw counts in your AnnData object and generate batch-corrected latent embeddings suitable for 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 perform batch correction on multi-study scRNA-seq datasets?▼

Batch correction on scRNA-seq datasets is performed using probabilistic variational models that learn batch-invariant latent representations by registering technical covariates like donor and condition directly within an AnnData object.

Can I use variational inference for differential expression in single-cell data?▼

Variational inference enables probabilistic differential expression in single-cell data by applying Bayesian posterior sampling to compute hypothesis testing modes such as change with configurable delta thresholds.

What is the best way to integrate multimodal CITE-seq and spatial transcriptomics data?▼

Integrating multimodal CITE-seq and spatial transcriptomics data is best handled by applying specialized probabilistic models like totalVI and DestVI to extract modality-specific latent representations from raw counts.

Do I need raw counts in AnnData to train single-cell variational models?▼

Raw counts in AnnData are required to train single-cell variational models, as the probabilistic workflows need count-based inputs alongside registered covariates to accurately estimate latent embeddings and normalized values.

Does scvi-tools support scATAC-seq and multiome integration workflows?▼

scATAC-seq and multiome integration workflows are supported through dedicated probabilistic models like MultiVI, which compute joint latent representations for integrated multi-omic analysis and differential accessibility testing.

When should I use probabilistic models for single-cell representation learning?▼

Probabilistic models for single-cell representation learning should be used when principled uncertainty handling is needed for dimensionality reduction, clustering, and extracting batch-corrected latent embeddings from complex multi-modal datasets.

Related Skills

Explore More Data & Analytics Skills→
Victory-HugoVictory-Hugo
1

scvi-tools

Run scvi-tools deep generative models on single-cell RNA sequencing data.

Community
Advanced
shushuznshushuzn
4

scvi-tools

Train probabilistic generative models on AnnData counts for batch-corrected embeddings.

Community
Advanced
jackspacejackspace
16

scvi-tools

Model single-cell omics data with scvi-tools for batch-corrected analysis.

Community
Advanced
Yezez9Yezez9

scvi-tools

Train scvi-tools models on AnnData objects for single-cell omics analysis.

Community
Advanced