gelidonyAMR

Oxford Nanopore-based Salmonella Infantis Genomic Epidemiology Pipeline

---

The name GelidonyAMR is inspired by the Gelidonya Lighthouse, a historic maritime landmark on the southern coast of Türkiye, near Finike. Just as the lighthouse has guided sailors for centuries, GelidonyAMR serves as a bioinformatics beacon for detecting antimicrobial resistance genes in Salmonella Infantis and other foodborne pathogens.

PhD Project: "Determination of Evolutionary Structure and Antimicrobial Resistance Profile of Salmonella Infantis"

---

Pipeline Overview

INPUT  (SRA accession list  OR  local FASTQ folder)
        │
        ▼
┌───────────────────────────────────┐
│  Quality Control                  │
│  FastQC · fastp · Kraken2         │
└────────────────┬──────────────────┘
                 │
                 ▼
┌───────────────────────────────────┐
│  Assembly                         │
│  Flye (Nano-raw)                  │
└────────────────┬──────────────────┘
                 │
                 ▼
┌───────────────────────────────────┐
│  Polishing & Assembly QC          │
│  Minimap2 + Racon (4x) · Clair3  │
│  QUAST · BUSCO · CheckM           │
└────────────────┬──────────────────┘
                 │
        ┌────────┴────────┐
        ▼                 ▼
┌──────────────┐  ┌────────────────┐
│  Annotation  │  │ Species Check  │
│  Prokka      │  │ FastANI        │
│  Bakta       │  └────────────────┘
└──────┬───────┘
       │
       ├── AMR Analysis ────── AMRFinderPlus · Abricate (7 databases)
       │
       ├── Plasmid Analysis ── PlasmidFinder · MobSuite
       │
       ├── Molecular Typing ── MLST · cgMLST · PopPUNK
       │
       ├── Phylogenomics ────── Parsnp · Roary → IQ-TREE
       │
       ├── Prophage ──────────── PHASTER
       │
       └── Variant Annotation ── Clair3 → snpEff

---

Requirements

• Operating System: Linux (Ubuntu 20.04+ recommended)
• Conda / Mamba: for environment management
• Nextflow: >= 23.10.0
• Java: 11 or higher (required by Nextflow)

Note: The PHASTER step requires an internet connection — it submits assemblies to the PHASTER web API (https://phaster.ca).

---

Installation

1. Create the Conda Environment

This pipeline includes many tools. Using mamba instead of standard conda is strongly recommended — it resolves dependencies 10-20x faster:

conda install -n base -c conda-forge mamba
mamba env create -f environment.yaml
conda activate gelidonyamr

Dependency note: chewBBACA, clair3, and poppunk can occasionally conflict with each other's Python requirements. If the environment fails to solve, try installing these in a separate environment and use Nextflow's conda process directive to assign them individually.

2. Set Up Databases

Kraken2

kraken2-build --standard --db /home/analysis/kraken2_db --threads 8

Abricate

abricate --setupdb
abricate --list     # Verify installed databases

chewBBACA cgMLST Schema

Downloaded automatically at runtime. To download manually:

chewBBACA.py DownloadSchema -sp 8 -sc 1 -o data/cgmlst/

Bakta

bakta_db download --output /home/analysis/bakta_db --type full

PlasmidFinder

git clone https://bitbucket.org/genomicepidemiology/plasmidfinder_db.git /home/analysis/plasmidfinder_db

snpEff — Salmonella Database

snpEff download Salmonella_enterica

SRA Toolkit Configuration (optional — improves download speed)

vdb-config --interactive
# Configure cache directory and cloud endpoint

---

Running the Pipeline

Mode 1 — Local FASTQ Folder

Place your Nanopore FASTQ files in a folder:

input_folder/
├── sample1.fastq
├── sample2.fastq
└── sample3.fastq

nextflow run main.nf \
  --reads "input_folder/*.fastq" \
  --outdir Results \
  -c config/nextflow.config

Mode 2 — NCBI SRA Accession List

Create a plain text file with one accession per line:

# sra_accessions.txt
SRR12345678
SRR12345679
SRR12345680
ERR9876543

• Accepts SRR, ERR, and DRR prefixes
• Lines starting with # are treated as comments and ignored
• Blank lines are skipped automatically

nextflow run main.nf \
  --sra_list sra_accessions.txt \
  --outdir Results \
  -c config/nextflow.config

--reads and --sra_list cannot be used together. If --sra_list is provided, --reads is ignored.

Additional Run Options

# Resume a pipeline that was interrupted
nextflow run main.nf -resume --reads "input_folder/*.fastq" -c config/nextflow.config

# Generate HTML execution report and timeline
nextflow run main.nf --reads "input_folder/*.fastq" -c config/nextflow.config \
  -with-report pipeline_report.html \
  -with-timeline pipeline_timeline.html

---

Parameters

Parameter	Default	Description
--reads	input_folder/*.fastq	Path to FASTQ files (glob patterns supported)
--sra_list	null	Path to TXT file containing SRA accession numbers
--outdir	Results	Directory where all results will be written
--ref_genome	data/ref_genome.fna	Reference genome (used by Clair3, FastANI, Parsnp)
--kraken2_db	/home/analysis/kraken2_db	Kraken2 database directory
--clair3_model	/opt/models/r941_prom_hac_g360+g422	Clair3 ONT chemistry model path
--bakta_db	/home/analysis/bakta_db	Bakta database directory
--plasmidfinder_db	/home/analysis/plasmidfinder_db	PlasmidFinder database directory
--snpeff_db	Salmonella_enterica	snpEff organism database name
--max_cpu	14	Maximum CPU threads
--max_memory	14GB	Maximum memory
--max_time	24h	Maximum run time

---

Output Directory Structure

Results/
├── raw_reads/          # FASTQ files downloaded from SRA (SRA mode only)
├── fastqc/             # Per-sample FastQC HTML reports
├── trimmed/            # fastp-trimmed reads
├── kraken2/            # Taxonomic classification reports
├── assembly/           # Raw Flye assembly output
├── polished/           # Final polished genomes (Minimap2 + Racon 4x)
├── quast/              # Assembly quality metrics (N50, contigs, etc.)
├── busco/              # Genome completeness scores
├── checkm/             # Contamination and completeness assessment
├── fastani/            # Species verification (ANI values vs. reference)
├── annotation/         # Prokka output (.gff, .gbk, .faa, .ffn)
├── bakta/              # Bakta annotation output
├── amr/                # Abricate results across 7 databases
├── amrfinder/          # NCBI AMRFinder Plus results
├── mobsuite/           # Plasmid reconstruction and mobility typing
├── plasmidfinder/      # Replicon-based plasmid typing
├── mlst/               # 7-gene MLST profiles
├── cgmlst/             # chewBBACA cgMLST / wgMLST allele calls
├── clair3/             # Variant calls (VCF)
├── snpeff/             # Annotated VCF files
├── phaster/            # Prophage detection results (JSON)
├── roary/              # Pan-genome analysis (core / accessory genes)
├── iqtree/             # Maximum likelihood phylogenetic tree
├── parsnp/             # Core SNP alignment and tree
├── poppunk/            # Population structure clusters
├── multiqc/            # Aggregated QC report across all samples
└── pipeline_info/      # Nextflow execution timeline and report

---

Tool Reference

Step	Tool	Version
SRA Download	SRA Toolkit (fasterq-dump)	≥3.0
Quality Control	FastQC	0.12.1
Read Trimming	fastp	0.24.0
Taxonomic Classification	Kraken2	2.1.3
Genome Assembly	Flye	2.9
Polishing	Minimap2 + Racon	≥2.26 / ≥1.5
Variant Calling	Clair3	≥1.0
Assembly Quality	QUAST	5.3.0
Genome Completeness	BUSCO	5.7.1
Quality Assessment	CheckM	≥1.2
Species Verification	FastANI	≥1.34
Annotation	Prokka + Bakta	≥1.14 / ≥1.9
AMR Detection	AMRFinder Plus	≥3.12
AMR Detection	Abricate	1.0.1
Plasmid Typing	PlasmidFinder	≥2.1
Plasmid Reconstruction	MobSuite	≥3.1
MLST	mlst	2.23.0
cgMLST / wgMLST	chewBBACA	3.3.6
Prophage Detection	PHASTER	Web API
Variant Annotation	snpEff	≥5.2
Pan-Genome	Roary	≥3.13
ML Phylogeny	IQ-TREE2	≥2.3
Core SNP Phylogeny	Parsnp	≥2.0
Population Structure	PopPUNK	≥2.6
QC Aggregation	MultiQC	1.27

---

Reference Genome

• Organism: Salmonella enterica subsp. enterica serovar Infantis
• Accession: LN649235.1
• NCBI Assembly: GCA_000953495.1

Download the reference genome:

datasets download genome accession GCA_000953495.1 --include genome

---

Frequently Asked Questions

Q: Do I need to configure anything extra for the SRA mode?

No. Simply pass --sra_list with the path to your accession file — the pipeline handles the rest. The only requirement is that sra-tools is installed (included in environment.yaml). Optionally, run vdb-config --interactive to configure AWS/GCP cloud endpoints for faster downloads.

Q: Do all tools run inside a single conda environment?

Most tools are available through the bioconda channel and can be installed via conda. However, chewBBACA, clair3, and poppunk can produce Python dependency conflicts. Use mamba instead of conda for significantly better dependency resolution. PHASTER requires no local installation — it runs entirely through the web API.

Q: The pipeline stopped midway. How do I resume it?

nextflow run main.nf -resume --reads "input_folder/*.fastq" -c config/nextflow.config

The -resume flag tells Nextflow to reuse cached results from completed steps.

Q: Can I run only specific steps?

Comment out the steps you don't need in main.nf using // and re-run. Nextflow's -resume flag ensures previously completed steps are not repeated.

---

Issues & Contact

For bug reports or feature requests, please open a GitHub issue.

---

Pipeline Developer: Gültekin Ünal

Inspired by the Gelidonya Lighthouse — Finike, Türkiye