Quality Control Pipelines at EGA

The European Genome-phenome Archive (EGA) provides automated quality control (QC) pipelines for common genomic file formats to ensure files are complete, well-structured, and ready for downstream use.

• The EGA currently stores a large number of phenotypic and genomic files — and it continues to grow thanks to the contributions of the scientific community.
• This repository includes Python-based pipelines for BAM/CRAM and VCF files, designed to extract key quality metrics and generate useful summaries.

---

Available QC Pipelines

• BAM/CRAM QC → Quality assessment for aligned read files (BAM/CRAM)
• VCF QC → Quality assessment for variant call files (VCF/VCF.GZ)

Each pipeline comes with example datasets, performance notes, and documentation.

---

BAM/CRAM QC

To improve the quality reports we generate for each of these files, we have developed a set of pipelines that automate the use of multiple bioinformatics tools for comprehensive quality assessment.

If you'd like to use these pipelines, please follow the Start Guide. For further details on how the scripts work, refer to the Documentation.

---

Example: BAM file from the 1000 Genomes Project

To illustrate the pipeline, we ran it on a small BAM file (Note: the BAM file contains alignments exclusively from chromosome 11) from the 1000 Genomes Project.

• Download the input file here
• Review the resulting output folder

It matches the structure and content you should expect if you follow the steps in the guide.

Learn more about the 1000 Genomes Project on their official website.

---

Running the BAM QC pipeline with Docker

1) Build the Docker image

From the root of the repository:

docker build -t bam-qc .

Important: Make sure the following files have execution permissions before building the image:

• run/qualimap_v2.3/qualimap
• run/BAM_pipeline_2.py
• output/BAM_finalize_2.py
• run/wrapper.py

If not, set them manually using:

chmod +x run/qualimap_v2.3/qualimap run/BAM_pipeline_2.py output/BAM_finalize_2.py run/wrapper.py

2) Run the pipeline

If your BAM, BED and FASTA files are located in /absolute/path/to/, run the container like this:

docker run --rm   -v /absolute/path/to:/data   -v $(pwd)/output:/app/output   bam-qc   --bam /data/muestra1.bam   --bed /data/regions.bed   --fasta /data/reference.fasta

This command:

• Mounts your local folder containing the input files as /data inside the container
• Mounts the repository's output/ directory (already created as BAM_QC/output/) to store the results
• Passes the required arguments (--bam, --bed, --fasta) to the pipeline
• Executes the full QC workflow and generates a multiqc_report.html in the output/ folder

3) Run MultiQC

To create the report please run inside the output/ folder:

multiqc . -e picard -e qualimap -c multiqc_config.yaml

You're done! Check the results in the multiqc_report.html file.

---

Benchmarking on other file types

We know the test file is relatively small, so we also evaluated the pipeline on:

• A 176 GB WGS BAM file
• A 5.8 GB RNA-seq BAM file

You can check the runtime performance and resource usage in the test/performance_logs folder.

---

VCF QC

The European Genome-phenome Archive (EGA) stores thousands of VCF files, which encode genetic variation data across samples and studies.

To ensure these files meet expected quality and formatting standards, we have developed a Python-based pipeline that performs initial QC checks and extracts summary metrics from VCF and VCF.GZ files.

If you'd like to use this pipeline, please follow the Start Guide. For detailed explanations of each step, refer to the Documentation.

---

Example: VCF QC on 1000 Genomes file

We also ran the VCF QC pipeline on a file from the 1000 Genomes Project:

• Download input VCF.
• View output folder.

Runtime and memory usage details are included in the corresponding performance log here.

🔗 VCF file format specification.

---

Together, these pipelines provide a consistent framework for checking the integrity and usability of genomic data hosted at EGA.