scRNA and snRNA-seq workflows with scverse tools (Scanpy + scvi-tools) integrated with PyTorch
Reproducible single-cell RNA-seq analysis using the scverse ecosystem
(Scanpy, scvi-tools, anndata) with PyTorch-backed variational models.
From raw counts to annotated clusters, differential expression, and clean figures.
Data Availability
Alzheimer's snRNA data available from: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE138852 (A single-cell atlas of entorhinal cortex from individuals with Alzheimer's disease reveals cell-type-specific gene expression regulation) (Grubman et al, 2019).
scRNA SARS-CoV-2 lung data from: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM5226574 A molecular single-cell lung atlas of lethal COVID-19 (Melms et al, 2021).
scRNA Immune Phenotype data from: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE114725 Single-cell Map of Diverse Immune Phenotypes in the Breast Tumor Microenvironment (Azizi et al 2019).
- QC & filtering: mitochondrial/ribosomal metrics, doublet flag support
- Normalization: log1p or scvi-tools normalized layers
- Batch integration: scVI latent space (
X_scVI) — runs on CPU or GPU via PyTorch - Clustering & UMAP: neighborhood graph, Leiden/UMAP
- Automated annotation: marker panels + per-cluster scoring
- Differential expression:
model.differential_expression(...)with FDR - Fast viz: UMAPs, dot/violin plots with grouped markers