scvi-tools

What problem does it solve?

This Skill helps you analyze single-cell omics data using scvi-tools, enabling probabilistic modeling, batch correction, and multimodal integration across modalities including scRNA-seq, scATAC-seq, CITE-seq, and spatial transcriptomics.

Core Features & Use Cases

• Probabilistic modeling: Build generative models (SCVI/autoencoders) for denoising, imputation, and latent representation learning.
• Batch correction and integration: Learn batch- or study-aware latent spaces for cross-dataset integration.
• Differential expression & annotation: Perform probabilistic differential expression and cell-type annotation with shared representations.
• Multimodal & spatial analysis: Integrate multimodal data and perform spatial analyses.
• Typical Workflow: Load data, configure SCVI, train, extract latent space, and perform downstream analyses.

Quick Start

Install scvi-tools, prepare an AnnData object, configure the SCVI model, train, and extract latent representations for downstream analysis.