genomics-pipeline

Step 18: Depth-Based CNV Calling with CNVnator

What This Does

Detects copy number variants (CNVs) using read-depth analysis — complementary to Manta’s paired-end/split-read approach. Especially effective for large CNVs (>1 kb) that Manta may miss.

Why

Manta (step 4) detects SVs from discordant read pairs and split reads, which works well for balanced SVs (inversions, translocations) and smaller deletions/duplications. CNVnator uses read-depth signal only, making it better for:

• Large duplications/deletions (>10 kb)
• Tandem duplications where breakpoints are ambiguous
• Validating Manta calls with orthogonal evidence

Tool

• CNVnator (Abyzov et al., Genome Research 2011)

Docker Image

quay.io/biocontainers/cnvnator:0.4.1--py312h99c8fb2_11

Command

# Step 1: Extract read mapping from BAM
docker run --rm \
  --cpus 4 --memory 8g \
  -v ${GENOME_DIR}:/genome \
  quay.io/biocontainers/cnvnator:0.4.1--py312h99c8fb2_11 \
  cnvnator \
    -root /genome/${SAMPLE}/cnvnator/${SAMPLE}.root \
    -tree /genome/${SAMPLE}/aligned/${SAMPLE}_sorted.bam

# Step 2: Generate read-depth histogram
docker run --rm \
  --cpus 4 --memory 8g \
  -v ${GENOME_DIR}:/genome \
  quay.io/biocontainers/cnvnator:0.4.1--py312h99c8fb2_11 \
  cnvnator \
    -root /genome/${SAMPLE}/cnvnator/${SAMPLE}.root \
    -his 1000 \
    -fasta /genome/reference/Homo_sapiens_assembly38.fasta

# Step 3: Statistics
docker run --rm \
  --cpus 4 --memory 8g \
  -v ${GENOME_DIR}:/genome \
  quay.io/biocontainers/cnvnator:0.4.1--py312h99c8fb2_11 \
  cnvnator \
    -root /genome/${SAMPLE}/cnvnator/${SAMPLE}.root \
    -stat 1000

# Step 4: Partition
docker run --rm \
  --cpus 4 --memory 8g \
  -v ${GENOME_DIR}:/genome \
  quay.io/biocontainers/cnvnator:0.4.1--py312h99c8fb2_11 \
  cnvnator \
    -root /genome/${SAMPLE}/cnvnator/${SAMPLE}.root \
    -partition 1000

# Step 5: Call CNVs
docker run --rm \
  --cpus 4 --memory 8g \
  -v ${GENOME_DIR}:/genome \
  quay.io/biocontainers/cnvnator:0.4.1--py312h99c8fb2_11 \
  cnvnator \
    -root /genome/${SAMPLE}/cnvnator/${SAMPLE}.root \
    -call 1000 \
    > ${GENOME_DIR}/${SAMPLE}/cnvnator/${SAMPLE}_cnvs.txt

Bin Size

The 1000 parameter is the bin size in base pairs. Use:

• 1000 for 30X WGS (recommended)
• 500 for higher coverage (>50X)
• 100 for targeted/exome data

Output

• ${SAMPLE}.root — ROOT file with read-depth data (intermediate, can be deleted)
• ${SAMPLE}_cnvs.txt — Tab-separated CNV calls with columns:
  • Type (deletion/duplication)
  • Coordinates (chr:start-end)
  • Size
  • Normalized read depth
  • e-value (statistical significance)
  • q0 (fraction of reads with mapping quality 0)

Filtering

# Keep only significant CNVs (e-value < 0.01, size > 1kb)
awk '$5 < 0.01 && $3 > 1000' ${SAMPLE}_cnvs.txt

Runtime

~2-4 hours per 30X WGS genome.

Notes

• Run AFTER alignment (step 2). Independent of Manta — can run in parallel.
• CNVnator and Manta overlap on ~60-70% of true CNVs. Variants called by both have lower false-positive rates than single-caller calls.
• The ROOT file format is from CERN’s ROOT framework — the Docker image includes all dependencies.
• For maximum sensitivity, intersect CNVnator calls with Manta and Delly (step 19) for a consensus call set.

This site is open source. Improve this page.