ONT-SV-CNV

Snakemake pipelines for structural variant (SV) and copy number variation (CNV) analysis from Oxford Nanopore long-read sequencing data.

Developed for multimodal sarcoma diagnostics combining nanopore methylation profiling with genomic structural analysis.

Pipeline overview

SV pipeline — Delly (long-read mode) → filtering (BND/TRA/INV/DUP) → GENCODE annotation → circos plot

CNV pipeline — mosdepth (coverage) → Spectre (CNV calling) → GENCODE annotation → genome-wide & per-chromosome plots

Repository structure

ONT-SV-CNV/
├── environment_linux.yml          # Conda environment (Linux)
├── environment_macos.yml          # Conda environment (macOS)
├── data/
│   ├── example_sample/            # Example data
│   │   ├── example_sample.bam
│   │   └── example_sample.bam.bai
│   └── reference/                 # User-provided reference files
│       ├── hg19.fasta
│       ├── hg19.fasta.fai
│       ├── gencode.v19.annotation.gtf
│       ├── hg19_blacklist.bed     # Optional
│       └── hg19.mdr               # Optional (auto-generated)
├── workflow/
│   ├── Analysis_ONT_SV            # SV Snakefile
│   ├── Analysis_ONT_CNV           # CNV Snakefile
│   └── scripts/
│       ├── annotate_BND.R
│       ├── annotate_CNV.R
│       ├── plot_circos.R
│       ├── plot_CNV.R
│       └── plot_CNV_chromosome.R
└── Output/                        # Generated automatically

---

Quick start

1. Clone the repository

git clone https://github.com/YOUR_USERNAME/ONT-SV-CNV.git
cd ONT-SV-CNV

2. Create the conda environment

Linux:

conda env create -f environment_linux.yml
conda activate ont_sv_cnv

macOS (Intel):

conda env create -f environment_macos.yml
conda activate ont_sv_cnv

macOS (Apple Silicon M1/M2/M3/M4):

CONDA_SUBDIR=osx-64 conda env create -f environment_macos.yml
conda activate ont_sv_cnv
conda config --env --set subdir osx-64

3. Download reference files

hg19 (default):

mkdir -p data/reference

# hg19 reference genome
wget -O data/reference/hg19.fasta.gz \
    https://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/hg19.fa.gz
gunzip data/reference/hg19.fasta.gz
samtools faidx data/reference/hg19.fasta

# GENCODE v19 annotation (hg19)
wget -O data/reference/gencode.v19.annotation.gtf.gz \
    https://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_19/gencode.v19.annotation.gtf.gz
gunzip data/reference/gencode.v19.annotation.gtf.gz

hg38:

mkdir -p data/reference

# hg38 reference genome (NCBI mirror — faster download)
wget -O data/reference/hg38.fasta.gz \
    https://ftp.ncbi.nlm.nih.gov/genomes/all/GCA/000/001/405/GCA_000001405.15_GRCh38/seqs_for_alignment_pipelines.ucsc_ids/GCA_000001405.15_GRCh38_no_alt_analysis_set.fna.gz
gunzip data/reference/hg38.fasta.gz
samtools faidx data/reference/hg38.fasta

# GENCODE v46 annotation (hg38)
wget -O data/reference/gencode.v46.annotation.gtf.gz \
    https://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_46/gencode.v46.annotation.gtf.gz
gunzip data/reference/gencode.v46.annotation.gtf.gz

Note: If using hg38, update the REF and GTF paths at the top of the Snakefiles accordingly.

4. Prepare your data

Place your BAM file (sorted, indexed) as:

data/example/example.bam
data/example/example.bam.bai

5. Run the pipelines

conda activate ont_sv_cnv

# SV analysis
snakemake -s workflow/Analysis_ONT_SV --cores 8 \
    --config sample="example"

# CNV analysis
snakemake -s workflow/Analysis_ONT_CNV --cores 8 \
    --config sample="example"

---

Parameters

SV pipeline

Parameter	Description	Default
sample	Required. Sample name (must match folder name in data/)	—
highlight_chr1	First chromosome to highlight in circos	NA
highlight_chr2	Second chromosome to highlight in circos	NA
sample_label	Display label for the circos plot	sample name

# Example with highlighting
snakemake -s workflow/Analysis_ONT_SV --cores 8 \
    --config sample="example" highlight_chr1="18" highlight_chr2="X" sample_label="Patient_01"

CNV pipeline

Parameter	Description	Default
sample	Required. Sample name (must match folder name in data/)	—
highlight_chr	Chromosome to highlight on genome-wide plot	NA
bin_size	Mosdepth coverage bin size (bp)	1000
min_cnv_len	Minimum CNV length for Spectre (bp)	100000
cancer	Enable Spectre cancer mode	false

# Example with cancer mode
snakemake -s workflow/Analysis_ONT_CNV --cores 8 \
    --config sample="example" highlight_chr="12" cancer=true

---

Output

SV results (Output/SV_Results/{sample}/)

File	Description
{sample}_delly.bcf	Raw Delly SV calls
{sample}_delly.vcf	Converted VCF
{sample}_delly_filtered.vcf	Filtered VCF (BND, TRA, INV, DUP only)
{sample}_BND_annotated.csv	Annotated breakpoints with gene names
figures/{sample}_circos.pdf/png	Circos plot

CNV results (Output/CNV_Results/{sample}/)

File	Description
{sample}.regions.bed.gz	Mosdepth coverage bins
{sample}.vcf.gz	Spectre CNV calls
{sample}_CNV_annotated.csv	Annotated CNVs with overlapping genes
figures/{sample}_CNV_genomewide.pdf/png	Genome-wide log2 ratio plot
figures/{sample}_CNV_{chr}.pdf/png	Per-chromosome plots (chr1–chrX)

---

Troubleshooting

• command not found — Make sure the environment is activated: conda activate ont_sv_cnv
• Apple Silicon install fails — Use the CONDA_SUBDIR=osx-64 prefix (see step 2)
• BAM not found — Check that your file is at data/SAMPLE/SAMPLE.bam with its .bam.bai index
• Re-run after failure — Snakemake resumes automatically; use --forceall for a clean re-run
• Dry run — Preview without executing: add -n -p to any snakemake command

---

Citation

If you use this pipeline, please cite:

• Delly: Rausch T, et al. Bioinformatics. 2012;28(18):i333–i339github.com/dellytools/delly.
• mosdepth: Pedersen BS, Quinlan AR. Bioinformatics. 2018;34(5):867–868github.com/brentp/mosdepth.

License

This project is licensed under the MIT License.