A framework for peptide identification on nanopore sequencing platforms

Nanopore peptide analysis workflow using Snakemake for preprocessing, signal segmentation, classification, and evaluation.

---

Pre-print:

• Beslic D, Kucklick M, Graap E, Sedaghatjoo S, Renard BY, Fuchs S, Engelmann S, Koerber N. A framework for peptide identification on commercial nanopore sequencing platforms. bioRxiv. 2026. https://doi.org/10.64898/2026.05.19.726067

---

Overview

This workflow performs:

1. POD5 preprocessing and Dorado basecalling
2. Read alignment to template references
3. Signal-level peptide segmentation
4. Feature extraction and classification
5. Evaluation and figure generation

The pipeline supports both local and SLURM-based execution.

---

Repository Structure

.
├── config/                 # Workflow configuration
├── data/                   # Input sequencing runs
├── Figures/                # Figure generation scripts
├── results/                # Processed outputs and intermediates
├── workflow/
│   ├── Snakefile           # Main workflow entrypoint
│   ├── rules/              # Snakemake rule definitions
│   └── scripts/            # Analysis scripts
├── environment.yml         # Conda/Mamba environment
├── local.sh                # Local workflow execution
├── slurm.sh                # SLURM workflow submission
└── README.md

---

Installation

Create the environment using conda:

conda env create -p ./env -f environment.yml
conda activate ./env

External Dependencies

Dorado is not installed automatically through the Conda environment and must be downloaded separately from Oxford Nanopore Technologies:

• https://github.com/nanoporetech/dorado

After installation, specify the Dorado executable path in:

config/config.yml

---

Input Data Structure

Place sequencing runs under:

data/<run_name>/pod5/

Example:

data/run01/pod5/

---

Configuration

Main configuration file:

config/config.yml

At minimum, verify:

dorado_path:
dorado_model:
ref_fasta:

---

Running the Workflow

Local execution

bash local.sh

SLURM execution

sbatch slurm.sh

---

Figure Reproduction

The repository includes processed serialized .pkl files required to recreate the manuscript figures.

Figure 2

python Figures/Figure2.py \
  --data-dir results/classification/features_dtw/ßCAT_single_variants/ \
  --output Figures/Figure2

Figure 3

python Figures/Figure3.py \
  --data-dir results/classification/inceptiontime/ \
  --output Figures/Figure3 

Figure 4

python Figures/Figure4.py \
  --data-dir results/classification/features_minirocket/ßCAT_single_variants/featuresLGBM/ \
  --output Figures/Figure4

---

Reproducibility Notes

Basecalling with Dorado may produce slight run-to-run differences depending on:

• GPU architecture
• CUDA version
• Number of GPUs used
• Parallelization settings

Such variability has been reported in the Dorado issue tracker (nanoporetech/dorado#617).

These effects may slightly affect individual read-level outputs and downstream counts but do not affect the conclusions of the study.

To ensure exact recreation of the published figures and reported metrics, this repository includes:

• serialized .pkl objects used for figure generation
• figure recreation scripts

Re-running the workflow from raw POD5 data may therefore yield slightly different numerical values, while preserving overall consistency of the results.

---

Tested Environment

Software	Version
Python	3.11.13
Snakemake	9.3.2
Dorado	1.3.1
CUDA	12.5

Tested on NVIDIA A100, H100, H200, and L40 GPUs.

---

Data Availability

Zenodo repository:

• https://doi.org/10.5281/zenodo.20269593

Contents include:

• raw POD5 files
• processed FASTQ files
• alignment outputs
• classification results