scientific-scvi-integration

What problem does it solve?

This pipeline enables end-to-end integration and probabilistic modeling for multi-batch single-cell data using scVI-tools, SCVI/SCANVI/totalVI and SOLO to unify datasets and extract meaningful latent representations.

Core Features & Use Cases

• Batch-aware integration: batch-corrected latent spaces for scRNA-seq across studies.
• Semi-supervised annotation: transfer cell type labels with high confidence using scANVI.
• Multi-modality support: combine RNA with protein (ADT) data via totalVI and detect doublets with SOLO.
• End-to-end pipeline: from data preparation to latent space visualization and clustering.

Quick Start

Run the scVI integration pipeline on your AnnData to generate a unified latent space and annotated cell types.