NanoFrag

Analyze DNA fragmentation and other patterns from liquid biopsy using nanopore sequencing.

---

Features

• Fragmentation Analysis: fragment size histogram and ratios.
• Copy Number Alterations: CNA detection using ichorCNA.
• Methylation Analysis: Get methylation patterns genome-wide, perform deconvolution into cell/tissues.
• Small Variant Detection: detect SNVs (not ready", experimental)

---

Installation

To clone the repository:

git clone https://github.com/bdolmo/nanofrag.git
cd nanofrag

Install dependencies:

• python 3.10+ is required.
• Install python dependencies with:

pip install -r requirements.txt

• Docker is required to run specific tools such as ichorCNV and nanomix

---

Docker Images

• seqeralabs/ichorcna:latest: Used for CNV analysis.
• bdolmo/nanomix:1.0.0: Used for methylation and fragmentation analysis.

You can pull them using the following command:

docker pull seqeralabs/ichorcna:latest
docker pull bdolmo/nanomix:1.0.0

---

Usage

python nanofrag.py --tumor_list <tumor_bam_list> --normal_list <normal_bam_list> \
                   --reference <reference_genome> --output_dir <output_dir> \
                   --threads <num_threads> [optional flags]

Required arguments

• --tumor_list: Path to a text file containing tumor BAM file paths (one per line).
• --normal_list: Path to a text file containing normal BAM file paths (one per line).
• --reference: Reference genome in FASTA format.
• --output_dir: Directory for output files.
• --threads: number of threads to use.

Optional flags (if you wish to skip any step)

• --skip_fragmentation: Skip fragmentation analysis.
• --skip_cn: Skip copy number analysis.
• --skip_methylation: Skip methylation analysis.
• --skip_small_variants: Skip small variant analysis (not ready)

---

Example

python nanofrag.py --tumor_list tumor_samples.txt \
                   --normal_list normal_samples.txt \
                   --reference hg38.fasta \
                   --output_dir results/ \
                   --threads 8

---

Important!

NanoFrag assumes that all bam files have been generated with methylation compatible tags (MM, and ML).

Output

The pipeline produces output files in the specified output_dir:

• Fragmentation Analysis:
  • Fragment size histograms and metrics.
• Copy Number Analysis:
  • CNV profiles in standard formats (e.g., .seg or .bed).
• Methylation Analysis:
  • Methylation profiles in .bed, deconvolution results in .txt.
• Small Variant Detection: (not ready)
  • VCF files with identified SNVs.

---