single-cell-rna-qc

Perform scRNA-seq quality control by calculating QC metrics on an AnnData object and applying MAD-based outlier detection to filter low-quality cells based on total counts, gene counts, and mitochondrial content.

MAD-based outlier detection for single-cell RNA-seq data identifies low-quality cells by flagging deviations in total counts, gene counts, and mitochondrial content using median absolute deviation, preserving high-quality cells for downstream analysis.

Generate QC visualizations before and after filtering in scanpy by running an end-to-end qc workflow that calculates metrics, applies cell and gene filters, and uses matplotlib to plot the before-and-after filtering results.

Yes, you can use this scRNA-seq QC workflow with .h5 file inputs, as it fully supports loading and processing both standard .h5ad and .h5 file formats to generate a clean AnnData object.

Yes, this single-cell quality control approach follows scverse best practices, automating standard preprocessing QC, MAD-based outlier detection, and hard MT% thresholds using scanpy and anndata dependencies.

Apply hard MT% thresholds when filtering scRNA-seq cells to enforce strict limits on mitochondrial gene expression, complementing MAD-based outlier detection to effectively remove dead or dying cells from the dataset.