Tumor-Normal Mutation Analysis Pipeline Using GATK Mutect2

I recently completed a project involving tumor-normal mutation analysis, and I wanted to share my experience with using the GATK Mutect2 pipeline for somatic mutation detection. This was a fantastic learning opportunity that allowed me to explore the intricacies of cancer genomics and the power of next-generation sequencing (NGS). Through this process, I gained a better understanding of how to identify somatic mutations like single nucleotide variants (SNVs) and small insertions/deletions (indels) by comparing tumor and normal samples.

Working through the GATK Mutect2 pipeline, I learned a lot about the importance of each step, from the initial quality control of raw reads to the final downstream analysis of mutations. I also discovered how critical it is to ensure high-quality data, from sequencing all the way through to variant calling and filtering. Below, I’ll walk you through the steps I took in the pipeline, share some of the tools I used, and highlight the key lessons I learned along the way.

Check out the full pipeline at > Github Repository

Introduction

Next-generation sequencing (NGS) has revolutionized cancer genomics, enabling precise identification of somatic mutations in tumor samples compared to their matched normal counterparts. The GATK Mutect2 pipeline is a powerful tool for detecting somatic single nucleotide variants (SNVs) and small insertions/deletions (indels) by comparing tumor and normal samples. This pipeline follows GATK Best Practices for somatic mutation calling and ensures high-confidence variant detection.

Pipeline Overview

The Tumor-Normal Mutation Analysis Pipeline follows these key steps:

• Quality Control (QC) of Raw Reads
• Alignment to the Reference Genome
• Post-Alignment Processing (Mark Duplicates, Base Quality Score Recalibration)
• Variant Calling with Mutect2
• Filtering and Annotation of Variants
• Downstream Analysis and Interpretation

Dependencies

• GATK >= 4.0
• BWA-MEM
• Samtools
• Picard
• FastQC
• bcftools
• Funcotator

Step 1: Quality Control of Raw Reads Before proceeding with analysis, it is crucial to check the quality of raw sequencing reads. This step ensures that low-quality bases and adapter sequences are removed, reducing errors in downstream analyses.

Tools Used:

• FastQC (for quality assessment)
• Trimmomatic (for adapter removal and trimming low-quality bases)

Command:

fastqc -o qc_reports/ sample_tumor_R1.fastq.gz sample_tumor_R2.fastq.gz
fastqc -o qc_reports/ sample_normal_R1.fastq.gz sample_normal_R2.fastq.gz

Step 2: Alignment to the Reference Genome High-quality reads are aligned to the human reference genome (hg38) using BWA-MEM, which ensures accurate mapping of reads.

Tools Used:

• BWA-MEM (Burrows-Wheeler Aligner)
• Samtools (for BAM file sorting and indexing)

Command:

bwa mem -t 8 -R "@RG\tID:sample\tSM:sample_tumor\tPL:ILLUMINA" hg38.fa sample_tumor_R1.fastq.gz sample_tumor_R2.fastq.gz | samtools sort -o sample_tumor_sorted.bam
bwa mem -t 8 -R "@RG\tID:sample\tSM:sample_normal\tPL:ILLUMINA" hg38.fa sample_normal_R1.fastq.gz sample_normal_R2.fastq.gz | samtools sort -o sample_normal_sorted.bam

Step 3: Post-Alignment Processing Post-alignment processing ensures that sequencing artifacts are minimized and the data is optimized for variant calling.

Sub-steps: Mark Duplicates

gatk MarkDuplicates -I sample_tumor_sorted.bam -O sample_tumor_dedup.bam -M tumor_metrics.txt

Base Quality Score Recalibration (BQSR)

gatk BaseRecalibrator -I sample_tumor_dedup.bam --known-sites known_variants.vcf -O recal_data.table
gatk ApplyBQSR -I sample_tumor_dedup.bam -bqsr recal_data.table -O sample_tumor_bqsr.bam

Step 4: Variant Calling with Mutect2 Mutect2 is used to identify somatic mutations by comparing tumor and normal samples.

Command:

gatk Mutect2 -R hg38.fa -I sample_tumor_bqsr.bam -I sample_normal_bqsr.bam -normal normal_sample -pon panel_of_normals.vcf -O raw_variants.vcf

Step 5: Filtering and Annotation To ensure high-confidence variants, filtering is applied using GATK’s FilterMutectCalls.

Command:

gatk FilterMutectCalls -V raw_variants.vcf -O filtered_variants.vcf

Further annotation of variants is done using ANNOVAR or VEP to gain insights into their biological significance.

Step 6: Downstream Analysis After obtaining filtered variants, further analysis can include:

• Functional annotation of variants
• Mutation burden analysis
• Pathway enrichment analysis

Conclusion

This Tumor-Normal Mutation Analysis Pipeline provides a structured and automated workflow for identifying somatic mutations in cancer genomics. By leveraging GATK Mutect2, this pipeline ensures high-quality, reliable variant detection, aiding researchers and clinicians in understanding tumor genetics and identifying potential therapeutic targets.

References

• https://www.rpubs.com/nnthieu/1241373
• https://nci-iteb.github.io/tumor_epidemiology_approaches/sessions/session_4/practical#somatic-short-variant-discovery-using-mutect2
• https://bioconductor.org/packages/devel/bioc/vignettes/maftools/inst/doc/maftools.html
• https://www.clinbioinfosspa.es/content/somatic-variant-pipeline