What models are supported by the scvi-tools skill?

It supports a wide range of models including scVI for RNA-seq, totalVI for CITE-seq, PeakVI for ATAC-seq, MultiVI for multiome, and DestVI for spatial data.

Can I use this skill for batch correction?

Yes, scvi-tools is specifically designed for unsupervised and semi-supervised batch correction and data integration using variational autoencoders (VAEs).

Is GPU support included in the documentation?

The skill includes an environment setup guide to help configure GPU acceleration for training deep learning models efficiently.

Can I map new data to a reference atlas?

Yes, it supports label transfer and reference mapping using models like scANVI and the scArches architecture for query-to-reference integration.

Does the skill handle data preparation?

Absolutely. It provides scripts and guidance for QC, filtering, and highly variable gene (HVG) selection using scvi-tools best practices.

scvi-tools Single-Cell Analysis

Name: scvi-tools Single-Cell Analysis
Author: anthropics

byanthropics

•

147

Ciencia de Datos y ML

Accelerates single-cell genomics research using probabilistic deep learning models for data integration, multi-modal analysis, and reference mapping.

Acerca de

This skill equips Claude with specialized knowledge of the scvi-tools framework, the industry standard for applying deep learning to single-cell genomics. It provides comprehensive guidance for performing unsupervised and semi-supervised data integration, multi-modal CITE-seq and multiome analysis, ATAC-seq peak analysis, and spatial transcriptomics deconvolution. By leveraging pre-built CLI scripts and Python utilities, users can efficiently process raw counts into meaningful latent representations, perform batch correction, and conduct differential expression analysis within a robust, reproducible workflow.

Características Principales

Automated data preparation and validation for deep learning models
Comprehensive multi-modal integration for RNA, ATAC, and protein data
Modular CLI scripts for training, clustering, and differential expression
Specialized workflows for spatial transcriptomics and RNA velocity
Reference-based label transfer and query mapping using scArches
147 GitHub stars

Casos de Uso

Integrating large-scale scRNA-seq datasets across multiple batches or technologies
Jointly analyzing multi-modal data such as CITE-seq or 10x Multiome using probabilistic models
Mapping query datasets to established reference atlases for automated cell-type annotation

Acerca de

Características Principales

Automated data preparation and validation for deep learning models
Comprehensive multi-modal integration for RNA, ATAC, and protein data
Modular CLI scripts for training, clustering, and differential expression
Specialized workflows for spatial transcriptomics and RNA velocity
Reference-based label transfer and query mapping using scArches
147 GitHub stars

Casos de Uso

Integrating large-scale scRNA-seq datasets across multiple batches or technologies
Jointly analyzing multi-modal data such as CITE-seq or 10x Multiome using probabilistic models
Mapping query datasets to established reference atlases for automated cell-type annotation