7: Somatic Oncology & Cancer Genomics

Chapter 7: Somatic Oncology & Cancer Genomics

Cancer is fundamentally a disease of the genome driven by acquired somatic mutations. Somatic variant calling differs dramatically from germline discovery: tumor biopsies are heterogeneous mixtures of malignant clones, normal stroma, and infiltrating immune cells. Consequently, somatic variants often present at low Variant Allele Fractions (VAF < 5–10%), requiring specialized statistical models to distinguish true mutations from sequencing noise.

1. GATK4 Mutect2 Somatic Calling Pipeline

# 1. Estimate Read Orientation Artifacts (vital for FFPE archival tissues)
gatk CollectF1R2Counts \
    -R GRCh38.fasta \
    -I tumor.bam \
    -O tumor_f1r2.tar.gz

gatk LearnReadOrientationModel \
    -I tumor_f1r2.tar.gz \
    -O read_orientation_model.tar.gz

# 2. Run Mutect2 in Paired Tumor-Normal Mode
gatk Mutect2 \
    -R GRCh38.fasta \
    -I tumor.bam \
    -I matched_normal.bam \
    -normal NORMAL_SAMPLE_NAME \
    --panel-of-normals pon.vcf.gz \
    --germline-resource af-only-gnomad.hg38.vcf.gz \
    --f1r2-tar-gz tumor_f1r2.tar.gz \
    -O tumor_unfiltered.vcf.gz

# 3. Calculate Contamination (cross-sample barcode bleeding)
gatk GetPileupSummaries \
    -I tumor.bam \
    -V common_biallelic.vcf.gz \
    -L common_biallelic.vcf.gz \
    -O tumor_getpileup.table

gatk CalculateContamination \
    -I tumor_getpileup.table \
    -O contamination.table

# 4. Filter Somatic Calls
gatk FilterMutectCalls \
    -R GRCh38.fasta \
    -V tumor_unfiltered.vcf.gz \
    --contamination-table contamination.table \
    --ob-priors read_orientation_model.tar.gz \
    -O tumor_somatic_filtered.vcf.gz