scvi-tools-single-cell

What problem does it solve?

This Skill addresses the complexity of analyzing single-cell omic data by providing advanced probabilistic models for batch correction, cell type annotation, and differential expression analysis, enabling deeper biological insights from noisy, high-dimensional datasets.

Core Features & Use Cases

• Probabilistic Batch Correction: Integrates multiple scRNA-seq datasets while preserving biological variation (scVI).
• Semi-Supervised Cell Annotation: Assigns cell types to unlabeled cells using a partially labeled reference (scANVI).
• Multi-modal Data Integration: Models both RNA and protein expression from CITE-seq data (totalVI).
• Transfer Learning: Adapts pre-trained models to new datasets efficiently (scARCHES).
• Differential Expression: Performs statistically grounded DE analysis with uncertainty quantification.
• Use Case: Integrate 10 different scRNA-seq experiments from varying protocols, identify novel cell subtypes using scANVI, and perform robust differential expression analysis between these subtypes.

Quick Start

Use the scvi-tools-single-cell skill to integrate multiple scRNA-seq batches and visualize the results.