alterlab-scvi-tools

What problem does it solve?

Provides a unified probabilistic framework to analyze single-cell data across modalities, integrating batch correction, normalization, and uncertainty-aware DE.

Core Features & Use Cases

• End-to-end probabilistic modeling: train VAEs for scRNA-seq, ATAC, and multimodal data with batch correction and covariate handling.
• Differential expression & imputation: perform Bayesian DE analyses and cross-modality imputation.
• Domain-wide use cases: RNA-seq, ATAC, CITE-seq, multiome, and spatial transcriptomics analyses.

Quick Start

Load an AnnData with raw counts, configure scvi tools models, and run training to obtain latent representations and DE results.