Single-Cell RNA-seq QC
Automated quality control for single-cell RNA-seq data with filtering and visualization
Bad cells with high mitochondrial content or low gene counts will corrupt your entire downstream analysis if not filtered early. This playbook automates single-cell RNA-seq quality control — detecting doublets, filtering low-quality cells, and generating QC visualizations before you proceed to clustering.
Who it's for: bioinformaticians running QC on 10X Chromium single-cell datasets before analysis, wet-lab biologists validating sequencing quality of their single-cell experiments, genomics core facility staff processing scRNA-seq data for multiple research groups, immunologists ensuring clean cell populations before differential expression analysis, computational biology students learning standard scRNA-seq QC workflows
Example
"Run QC on our 10X scRNA-seq dataset and filter low-quality cells" → scRNA-seq QC pipeline: per-cell metric calculation (gene counts, UMI counts, mitochondrial percentage), QC violin plots and scatter visualizations for threshold selection, doublet detection and removal, adaptive filtering based on distribution-derived thresholds, and clean count matrix export ready for normalization and clustering
CLAUDE.md Template
New here? 3-minute setup guide → | Already set up? Copy the template below.
npx ccplaybooks add single-cell-rna-qc# Single-Cell RNA-seq Quality Control
Automated QC workflow for single-cell RNA-seq data following scverse best practices.
## When to Use This Playbook
Use when you need to:
- Perform quality control on single-cell RNA-seq data
- Filter low-quality cells or assess data quality
- Generate QC visualizations and metrics
- Follow scverse/scanpy best practices
- Apply MAD-based filtering or outlier detectionGet new playbooks like this one
One email a week with new Claude Code workflows. Free, like everything here.
No spam. Unsubscribe anytime.
What This Does
This playbook automates quality control for single-cell RNA-seq data following scverse best practices. It calculates QC metrics (count depth, gene detection, mitochondrial content), applies MAD-based outlier filtering, and generates comprehensive before/after visualizations -- giving you a clean, filtered dataset ready for downstream analysis like clustering and cell type annotation.
Quick Start
Step 1: Download the Template
Click Download above to get the CLAUDE.md file.
Step 2: Set Up Your Project
Create a project folder with your data and the template:
my-scrna-project/
CLAUDE.md
data/
sample.h5ad # or raw_feature_bc_matrix.h5
Step 3: Start Working
claude
Say: "Run QC on my single-cell RNA-seq data in data/sample.h5ad"
Two Approaches
Approach 1: Complete Pipeline (Recommended)
A one-command solution for standard QC workflows. Handles metrics calculation, MAD-based filtering, gene filtering, and visualization generation automatically.
Approach 2: Modular Building Blocks
For custom workflows where you need different thresholds for cell subsets, conditional filtering logic, or partial execution (metrics only, no filtering).
Output Files
The pipeline generates these files in the results directory:
qc_metrics_before_filtering.png-- Pre-filtering visualizationsqc_filtering_thresholds.png-- MAD-based threshold overlaysqc_metrics_after_filtering.png-- Post-filtering quality metrics*_filtered.h5ad-- Clean dataset ready for downstream analysis*_with_qc.h5ad-- Original data with QC annotations preserved
Tips
- Be permissive with filtering -- default thresholds retain most cells to avoid losing rare populations
- Always review before/after plots to ensure filtering makes biological sense
- Some tissues naturally have higher mitochondrial content (neurons, cardiomyocytes) -- adjust thresholds accordingly
- Mitochondrial gene prefixes vary by species:
mt-for mouse,MT-for human - QC parameters may need adjustment based on the specific experiment or tissue type
Supported Formats
| Format | Source | Extension |
|---|---|---|
| AnnData | scanpy/Python workflows | .h5ad |
| 10X Genomics | Cell Ranger output | .h5 |
Example Prompts
"Run QC on my single-cell data in sample.h5ad with default settings."
"Analyze the quality of my 10X Genomics h5 file and filter low-quality cells."
"Calculate QC metrics only -- don't filter yet. I want to inspect the distributions first."
"Run QC with a stricter mitochondrial threshold of 10% instead of the default."
"Apply different QC thresholds for neurons vs other cell types in my dataset."