MethylationArray-snake

Overview

This pipeline performs analysis of Illumina DNA methylation array data using a Snakemake workflow. It is designed for processing raw IDAT files from EPIC / Infinium methylation arrays.

The workflow combines:

• minfi for IDAT import, quality control, preprocessing and normalization
• Probe filtering for problematic, cross-reactive and SNP-associated probes
• Conversion to .h5ad format for downstream analysis in Python / Scanpy workflows
• Copy number analysis derived from methylation intensities
• Tumor purity estimation using methylation-based models

The main goal is to generate high-quality normalized methylation data together with copy number profiles and tumor purity estimates.

---

Workflow Summary

1. Sample Preparation

Input samplesheets are defined in config.yaml for each dataset. This script has to be edited by the user.

2. Preprocessing of IDAT Files

For each configured dataset:

• Raw IDAT files are imported
• Quality control is performed
• Signal intensities are normalized
• Problematic probes are removed
• Beta-values / M-values are computed
• Data are exported as R objects and .h5ad files

3. Copy Number Analysis

Using methylation intensity data:

• CNA profiles are inferred from array signal intensities
• The MINT cohort is analyzed using the Pai cohort as reference normals
• Segmented CNA profiles and plots are generated

4. Tumor Purity Estimation

Tumor purity is estimated using:

• RF_purity
• InfiniumPurify

5. Reporting / Export

Processed methylation objects, CNA profiles and purity tables are written to disk.

---

Software Requirements

• Snakemake 9.17.3
• Conda / Mamba

---

Running the Pipeline

snakemake --cores 1 --use-conda

---

Main Output Files

Normalized Methylation Data

methylation/{dataset}/methylation_data_{norm}.h5ad

Primary output containing processed methylation data for downstream analysis.

R Methylation Objects

methylation/{dataset}/methylation_object_{norm}.Rds

Contains normalized minfi objects for R-based downstream analyses.

Copy Number Output

CNAs/MINT/Segmented_CNAs_MINT.txt

Primary CNA segmentation output derived from methylation array signal intensities.

Tumor Purity Estimates

results/Tumor_purities.txt

Per-sample tumor purity estimates generated from methylation data using RFpurity and InfiniumPurify.

Additional Outputs

CNA Plots

CNAs/MINT/plots/

Contains per-sample CNA profile visualizations.

---

Directory Structure

output/
├── methylation/
├── CNAs/
└── results/

---

Notes on Illumina Array Processing

This pipeline supports Illumina methylation array workflows where preprocessing quality is essential.

By default, problematic probes can be removed using supplied reference lists:

• Zhou probes
• Cross-reactive probes
• Other problematic probes

Normalization methods are defined in config.yaml.

---