Pydeseq2

PyDESeq2

Overview

PyDESeq2 is a Python implementation of DESeq2 for differential expression analysis with bulk RNA-seq data. Design and execute complete workflows from data loading through result interpretation, including single-factor and multi-factor designs, Wald tests with multiple testing correction, optional apeGLM shrinkage, and integration with pandas and AnnData.

When to Use This Skill

This skill should be used when:

• Analyzing bulk RNA-seq count data for differential expression
• Comparing gene expression between experimental conditions (e.g., treated vs control)
• Performing multi-factor designs accounting for batch effects or covariates
• Converting R-based DESeq2 workflows to Python
• Integrating differential expression analysis into Python-based pipelines
• Users mention "DESeq2", "differential expression", "RNA-seq analysis", or "PyDESeq2"

Quick Start Workflow

For users who want to perform a standard differential expression analysis:

import pandas as pd
from pydeseq2.dds import DeseqDataSet
from pydeseq2.ds import DeseqStats

# 1. Load data
counts_df = pd.read_csv("counts.csv", index_col=0).T  # Transpose to samples × genes
metadata = pd.read_csv("metadata.csv", index_col=0)

# 2. Filter low-count genes
genes_to_keep = counts_df.columns[counts_df.sum(axis=0) >= 10]
counts_df = counts_df[genes_to_keep]

# 3. Initialize and fit DESeq2
dds = DeseqDataSet(
    counts=counts_df,
    metadata=metadata,
    design="~condition",
    refit_cooks=True
)
dds.deseq2()

# 4. Perform statistical testing
ds = DeseqStats(dds, contrast=["condition", "treated", "control"])
ds.summary()

# 5. Access results
results = ds.results_df
significant = results[results.padj = 10]
counts_df = counts_df[genes_to_keep]

# Remove samples with missing metadata
samples_to_keep = ~metadata.condition.isna()
counts_df = counts_df.loc[samples_to_keep]
metadata = metadata.loc[samples_to_keep]

Step 2: Design Specification

The design formula specifies how gene expression is modeled.

Single-factor designs:

design = "~condition"  # Simple two-group comparison

Multi-factor designs:

design = "~batch + condition"  # Control for batch effects
design = "~age + condition"     # Include continuous covariate
design = "~group + condition + group:condition"  # Interaction effects

Design formula guidelines:

• Use Wilkinson formula notation (R-style)
• Put adjustment variables (e.g., batch) before the main variable of interest
• Ensure variables exist as columns in the metadata DataFrame
• Use appropriate data types (categorical for discrete variables)

Step 3: DESeq2 Fitting

Initialize the DeseqDataSet and run the complete pipeline:

from pydeseq2.dds import DeseqDataSet

dds = DeseqDataSet(
    counts=counts_df,
    metadata=metadata,
    design="~condition",
    refit_cooks=True,  # Refit after removing outliers
    n_cpus=1           # Parallel processing (adjust as needed)
)

# Run the complete DESeq2 pipeline
dds.deseq2()

What deseq2() does:

1. Computes size factors (normalization)
2. Fits genewise dispersions
3. Fits dispersion trend curve
4. Computes dispersion priors
5. Fits MAP dispersions (shrinkage)
6. Fits log fold changes
7. Calculates Cook's distances (outlier detection)
8. Refits if outliers detected (optional)

Step 4: Statistical Testing

Perform Wald tests to identify differentially expressed genes:

from pydeseq2.ds import DeseqStats

ds = DeseqStats(
    dds,
    contrast=["condition", "treated", "control"],  # Test treated vs control
    alpha=0.05,                # Significance threshold
    cooks_filter=True,         # Filter outliers
    independent_filter=True    # Filter low-power tests
)

ds.summary()

Contrast specification:

• Format: [variable, test_level, reference_level]
• Example: ["condition", "treated", "control"] tests treated vs control
• If None, uses the last coefficient in the design

Result DataFrame columns:

• baseMean: Mean normalized count across samples
• log2FoldChange: Log2 fold change between conditions
• lfcSE: Standard error of LFC
• stat: Wald test statistic
• pvalue: Raw p-value
• padj: Adjusted p-value (FDR-corrected via Benjamini-Hochberg)

Step 5: Optional LFC Shrinkage

Apply shrinkage to reduce noise in fold change estimates:

ds.lfc_shrink()  # Applies apeGLM shrinkage

When to use LFC shrinkage:

• For visualization (volcano plots, heatmaps)
• For ranking genes by effect size
• When prioritizing genes for follow-up experiments

Important: Shrinkage affects only the log2FoldChange values, not the statistical test results (p-values remain unchanged). Use shrunk values for visualization but report unshrunken p-values for significance.

Step 6: Result Export

Save results and intermediate objects:

import pickle

# Export results as CSV
ds.results_df.to_csv("deseq2_results.csv")

# Save significant genes only
significant = ds.results_df[ds.results_df.padj  1)
]

# Separate up- and down-regulated
upregulated = significant[significant.log2FoldChange > 0]
downregulated = significant[significant.log2FoldChange  samples, transpose is needed
if counts_df.shape[1] < counts_df.shape[0]:
    counts_df = counts_df.T

Design Matrix Issues

Issue: "Design matrix is not full rank"

Cause: Confounded variables (e.g., all treated samples in one batch)

Solution: Remove confounded variable or add interaction term

# Check confounding
print(pd.crosstab(metadata.condition, metadata.batch))

# Either simplify design or add interaction
design = "~condition"  # Remove batch
# OR
design = "~condition + batch + condition:batch"  # Model interaction

No Significant Genes

Diagnostics:

# Check dispersion distribution
plt.hist(dds.varm["dispersions"], bins=50)
plt.show()

# Check size factors
print(dds.obsm["size_factors"])

# Look at top genes by raw p-value
print(ds.results_df.nsmallest(20, "pvalue"))

Possible causes:

• Small effect sizes
• High biological variability
• Insufficient sample size
• Technical issues (batch effects, outliers)

Reference Documentation

For comprehensive details beyond this workflow-oriented guide:

• API Reference (references/api_reference.md): Complete documentation of PyDESeq2 classes, methods, and data structures. Use when needing detailed parameter information or understanding object attributes.

• Workflow Guide (references/workflow_guide.md): In-depth guide covering complete analysis workflows, data loading patterns, multi-factor designs, troubleshooting, and best practices. Use when handling complex experimental designs or encountering issues.

Load these references into context when users need:

• Detailed API documentation: Read references/api_reference.md
• Comprehensive workflow examples: Read references/workflow_guide.md
• Troubleshooting guidance: Read references/workflow_guide.md (see Troubleshooting section)

Key Reminders

1. Data orientation matters: Count matrices typically load as genes × samples but need to be samples × genes. Always transpose with .T if needed.

1. Sample filtering: Remove samples with missing metadata before analysis to avoid errors.

1. Gene filtering: Filter low-count genes (e.g., < 10 total reads) to improve power and reduce computational time.

1. Design formula order: Put adjustment variables before the variable of interest (e.g., "~batch + condition" not "~condition + batch").

1. LFC shrinkage timing: Apply shrinkage after statistical testing and only for visualization/ranking purposes. P-values remain based on unshrunken estimates.

1. Result interpretation: Use padj < 0.05 for significance, not raw p-values. The Benjamini-Hochberg procedure controls false discovery rate.

1. Contrast specification: The format is [variable, test_level, reference_level] where testlevel is compared against referencelevel.

1. Save intermediate objects: Use pickle to save DeseqDataSet objects for later use or additional analyses without re-running the expensive fitting step.

Installation and Requirements

uv pip install pydeseq2

System requirements:

• Python 3.10-3.11
• pandas 1.4.3+
• numpy 1.23.0+
• scipy 1.11.0+
• scikit-learn 1.1.1+
• anndata 0.8.0+

Optional for visualization:

• matplotlib
• seaborn

Additional Resources

• Official Documentation: https://pydeseq2.readthedocs.io
• GitHub Repository: https://github.com/owkin/PyDESeq2
• Publication: Muzellec et al. (2023) Bioinformatics, DOI: 10.1093/bioinformatics/btad547
• Original DESeq2 (R): Love et al. (2014) Genome Biology, DOI: 10.1186/s13059-014-0550-8

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: LeonChaoX
• Source: LeonChaoX/qinyan-academic-skills
• License: MIT

Install and usage instructions live in the source repository linked above.