scanpy

End-to-end scRNA-seq analysis runs standard preprocessing steps from QC through normalization, HVG selection, and scaling, then generates UMAP embeddings and Leiden clusters to identify cell types. The workflow processes AnnData-based datasets to produce publication-ready figures and cluster annotations.

To identify marker genes for single-cell clusters, run Leiden clustering followed by the rank_genes_groups function to rank differentially expressed genes per cluster. This workflow computes annotations and maps clusters to cell types for downstream biological interpretation.

Yes, this scRNA-seq workflow supports 10x-derived inputs stored in AnnData-compatible formats. It computes mitochondrial QC metrics, executes standard preprocessing, and runs sc.tl and sc.pl functions including neighbors, UMAP, Leiden, and rank_genes_groups for visualization.

Yes, you can compute mitochondrial QC metrics and filter low-quality cells using this pipeline. The workflow visualizes QC distributions and applies filtering thresholds to remove low-quality cells and genes before proceeding to normalization and dimensionality reduction.

Before Leiden clustering on an AnnData dataset, required normalization steps include log transformation, highly variable gene selection, scaling, and regression. The workflow then computes PCA embeddings and neighbor graphs to enable the Leiden clustering algorithm.

An exploratory scRNA-seq pipeline is not suited for non-AnnData-compatible data or workflows requiring non-standard preprocessing outside of standard QC, normalization, and clustering steps. It is designed for exploratory analysis rather than specialized analyses deviating from standard sc.tl and sc.pl functions.