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 Skill

Use when users:

• Request quality control or QC on single-cell RNA-seq data
• Want to filter low-quality cells or assess data quality
• Need QC visualizations or metrics
• Ask to follow scverse/scanpy best practices
• Request MAD-based filtering or outlier detection

Supported input formats:

• .h5ad files (AnnData format from scanpy/Python workflows)
• .h5 files (10X Genomics Cell Ranger output)

Default recommendation: Use Approach 1 (complete pipeline) unless the user has specific custom requirements or explicitly requests non-standard filtering logic.

Approach 1: Complete QC Pipeline (Recommended for Standard Workflows)

For standard QC following scverse best practices, use the convenience script scripts/qc_analysis.py:

python3 scripts/qc_analysis.py input.h5ad
# or for 10X Genomics .h5 files:
python3 scripts/qc_analysis.py raw_feature_bc_matrix.h5

The script automatically detects the file format and loads it appropriately.

When to use this approach:

• Standard QC workflow with adjustable thresholds (all cells filtered the same way)
• Batch processing multiple datasets
• Quick exploratory analysis
• User wants the “just works” solution

Requirements: anndata, scanpy, scipy, matplotlib, seaborn, numpy

Parameters:

Customize filtering thresholds and gene patterns using command-line parameters:

• --output-dir - Output directory
• --mad-counts, --mad-genes, --mad-mt - MAD thresholds for counts/genes/MT%
• --mt-threshold - Hard mitochondrial % cutoff
• --min-cells - Gene filtering threshold
• --mt-pattern, --ribo-pattern, --hb-pattern - Gene name patterns for different species

Use --help to see current default values.

Outputs:

All files are saved to <input_basename>_qc_results/ directory by default (or to the directory specified by --output-dir):

• qc_metrics_before_filtering.png - Pre-filtering visualizations
• qc_filtering_thresholds.png - MAD-based threshold overlays
• qc_metrics_after_filtering.png - Post-filtering quality metrics
• <input_basename>_filtered.h5ad - Clean, filtered dataset ready for downstream analysis
• <input_basename>_with_qc.h5ad - Original data with QC annotations preserved

If copying outputs for user access, copy individual files (not the entire directory) so users can preview them directly.

Workflow Steps

The script performs the following steps:

1. Calculate QC metrics - Count depth, gene detection, mitochondrial/ribosomal/hemoglobin content
2. Apply MAD-based filtering - Permissive outlier detection using MAD thresholds for counts/genes/MT%
3. Filter genes - Remove genes detected in few cells
4. Generate visualizations - Comprehensive before/after plots with threshold overlays

Approach 2: Modular Building Blocks (For Custom Workflows)

For custom analysis workflows or non-standard requirements, use the modular utility functions from scripts/qc_core.py and scripts/qc_plotting.py:

# Run from scripts/ directory, or add scripts/ to sys.path if needed
import anndata as ad
from qc_core import calculate_qc_metrics, detect_outliers_mad, filter_cells
from qc_plotting import plot_qc_distributions  # Only if visualization needed

adata = ad.read_h5ad('input.h5ad')
calculate_qc_metrics(adata, inplace=True)
# ... custom analysis logic here

When to use this approach:

• Different workflow needed (skip steps, change order, apply different thresholds to subsets)
• Conditional logic (e.g., filter neurons differently than other cells)
• Partial execution (only metrics/visualization, no filtering)
• Integration with other analysis steps in a larger pipeline
• Custom filtering criteria beyond what command-line params support

Available utility functions:

From qc_core.py (core QC operations):

• calculate_qc_metrics(adata, mt_pattern, ribo_pattern, hb_pattern, inplace=True) - Calculate QC metrics and annotate adata
• detect_outliers_mad(adata, metric, n_mads, verbose=True) - MAD-based outlier detection, returns boolean mask
• apply_hard_threshold(adata, metric, threshold, operator='>', verbose=True) - Apply hard cutoffs, returns boolean mask
• filter_cells(adata, mask, inplace=False) - Apply boolean mask to filter cells
• filter_genes(adata, min_cells=20, min_counts=None, inplace=True) - Filter genes by detection
• print_qc_summary(adata, label='') - Print summary statistics

From qc_plotting.py (visualization):

• plot_qc_distributions(adata, output_path, title) - Generate comprehensive QC plots
• plot_filtering_thresholds(adata, outlier_masks, thresholds, output_path) - Visualize filtering thresholds
• plot_qc_after_filtering(adata, output_path) - Generate post-filtering plots

Example custom workflows:

Example 1: Only calculate metrics and visualize, don’t filter yet

adata = ad.read_h5ad('input.h5ad')
calculate_qc_metrics(adata, inplace=True)
plot_qc_distributions(adata, 'qc_before.png', title='Initial QC')
print_qc_summary(adata, label='Before filtering')

Example 2: Apply only MT% filtering, keep other metrics permissive

adata = ad.read_h5ad('input.h5ad')
calculate_qc_metrics(adata, inplace=True)

# Only filter high MT% cells
high_mt = apply_hard_threshold(adata, 'pct_counts_mt', 10, operator='>')
adata_filtered = filter_cells(adata, ~high_mt)
adata_filtered.write('filtered.h5ad')

Example 3: Different thresholds for different subsets

adata = ad.read_h5ad('input.h5ad')
calculate_qc_metrics(adata, inplace=True)

# Apply type-specific QC (assumes cell_type metadata exists)
neurons = adata.obs['cell_type'] == 'neuron'
other_cells = ~neurons

# Neurons tolerate higher MT%, other cells use stricter threshold
neuron_qc = apply_hard_threshold(adata[neurons], 'pct_counts_mt', 15, operator='>')
other_qc = apply_hard_threshold(adata[other_cells], 'pct_counts_mt', 8, operator='>')

Best Practices

1. Be permissive with filtering - Default thresholds intentionally retain most cells to avoid losing rare populations
2. Inspect visualizations - Always review before/after plots to ensure filtering makes biological sense
3. Consider dataset-specific factors - Some tissues naturally have higher mitochondrial content (e.g., neurons, cardiomyocytes)
4. Check gene annotations - Mitochondrial gene prefixes vary by species (mt- for mouse, MT- for human)
5. Iterate if needed - QC parameters may need adjustment based on the specific experiment or tissue type

Reference Materials

For detailed QC methodology, parameter rationale, and troubleshooting guidance, see references/scverse_qc_guidelines.md. This reference provides:

• Detailed explanations of each QC metric and why it matters
• Rationale for MAD-based thresholds and why they’re better than fixed cutoffs
• Guidelines for interpreting QC visualizations (histograms, violin plots, scatter plots)
• Species-specific considerations for gene annotations
• When and how to adjust filtering parameters
• Advanced QC considerations (ambient RNA correction, doublet detection)

Load this reference when users need deeper understanding of the methodology or when troubleshooting QC issues.

Next Steps After QC

Typical downstream analysis steps:

• Ambient RNA correction (SoupX, CellBender)
• Doublet detection (scDblFinder)
• Normalization (log-normalize, scran)
• Feature selection and dimensionality reduction
• Clustering and cell type annotation