Polio Whole-Capsid NGS Surveillance Workflow

The repository implements the Illumina whole-capsid approach after culture and provides comparable outputs to conventional GPLN.

Associated manuscript: Comparative Evaluation of Whole-Capsid Next-Generation Sequencing and Traditional Methods in Post-Culture Polio Surveillance: Pakistan and Afghanistan Study (2021-2024).

Abbreviations

Term	Meaning
NGS	Next-generation sequencing
VP1	Viral protein 1
VP2	Viral protein 2
VP3	Viral protein 3
FASTQ	Sequencing-read file format with base calls and quality scores
BAM	Binary Alignment/Map file for aligned sequencing reads
QC	Quality control
TSV	Tab-separated values table
MAFFT	Multiple-sequence alignment program
IQ-TREE	Maximum-likelihood phylogenetic tree program

Program summary

This repository contains one end to end pipeline that can reproduce the analyses reported in the manuscript. It supports the following components under one roof:

• Inputs
  • Paired end FASTQ from cultured isolates prepared for whole capsid sequencing on Illumina MiSeq 2x150.
• Quality control and trimming
  • fastp for adapter and quality trimming with configurable minimum base quality, default Q20, and minimum length 50.
• Mapping and primer trimming
  • BWA MEM mapping to references for WPV1 and Sabin strains.
  • BAMClipper to remove primer sequences when amplicon primers are used.
  • Picard MarkDuplicates to mark PCR duplicates.
• Coverage and consensus
  • BEDTools genomecov to compute depth profiles.
  • FreeBayes variant calling followed by BCFtools consensus with a depth mask. Default minimum depth 10.
• Antigenic site analysis
  • Translate VP1, VP2, VP3. Summarize amino acid changes in canonical antigenic sites Ag1, Ag2, Ag3. Output tidy TSV reports per sample.
• Phylogeny and context
  • Fetch selected context references from GenBank.
  • MAFFT alignment of masked consensuses plus context.
  • IQ-TREE ML tree with 1000 ultrafast bootstraps. ModelFinder can be toggled on.
• Statistics
  • Optional scripts to compute McNemar test and paired t test when you provide detection tables and per site identities from Sanger vs NGS.
• Optional de novo assembly
  • SPAdes assembly and contig QC, BLAST for sanity checking and for identifying best reference when needed.

The workflow is implemented in Snakemake and can be run on a laptop or a small server. No clinical data are included.

Current software version: 1.0.3. See CHANGELOG.md and CITATION.cff for release metadata.

Workflow overview

This repository presents an Illumina whole-capsid analysis path for poliovirus surveillance, from paired-end reads to masked consensus sequences, antigenic-site summaries, phylogeny, and optional statistical comparison with conventional methods. It is designed to be understandable to both bioinformaticians and public-health genomics supervisors.

flowchart LR
  A["Cultured isolate FASTQ"] --> B["QC and trimming"]
  B --> C["Reference mapping"]
  C --> D["Depth and variant review"]
  D --> E["Masked consensus"]
  E --> F["Antigenic-site table"]
  E --> G["Phylogeny"]
  F --> H["Public-health report"]
  G --> H

Loading

Requirements

• Linux, macOS, or WSL with Bash
• Python 3.10 or newer for lightweight script checks
• Conda or mamba for the full workflow environment
• Snakemake 7 or newer for the full pipeline
• System tools installed through conda: fastp, bwa, samtools, bcftools, freebayes, bedtools, picard, bamclipper, mafft, iqtree, spades, seqkit, entrez-direct

NCBI usage note

Set a contact email once per shell for E-utilities. Optional API key improves rate limits.

export NCBI_EMAIL="you@example.com"
export NCBI_API_KEY="xxxxxxxxxxxxxxxxxxxxxxxxxxxx"   # optional

Quick verification

This check validates the Python utilities and example plotting without requiring private FASTQ files.

python3 -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install -r env/requirements.txt
python -m compileall -q analysis/scripts
python analysis/scripts/example_qc_plot.py --in data-example/example_counts.tsv --out results-example/example_plot.png

The same check can be run with:

make test

Full workflow run

No raw FASTQ files are included in this repository. Before running the full workflow, edit config/config.yaml or pass a separate config file with paths to local paired-end FASTQ files.

export NCBI_EMAIL="you@example.com"
mamba env create -f env/environment.yml   # or: conda env create -f env/environment.yml
conda activate polio-capsid-env
snakemake -s workflow/Snakefile --configfile config/config.yaml -n -c 4
snakemake -s workflow/Snakefile --configfile config/config.yaml -c 4 --printshellcmds

The make wrappers accept THREADS, CONFIG, PYTHON, and SNAKEMAKE overrides:

make dry CONFIG=config/config.yaml THREADS=4
make run CONFIG=config/config.yaml THREADS=4

Configuration

Edit config/config.yaml. Minimal example:

pairs:
  - sample: PV1_001
    r1: data-private/PV1_001_R1.fastq.gz
    r2: data-private/PV1_001_R2.fastq.gz

references:
  wpv1: KY941935.1
  sabin1: V01150.1
  sabin2: AY184220.1
  sabin3: AY184221.1

context_acc:
  - KY941935.1
  - V01150.1
  - AY184220.1
  - AY184221.1

params:
  threads: 4
  min_len: 50
  min_qual: 20
  min_depth_consensus: 10
  use_model_finder: false
  iqtree_model: GTR+G+I
  bootstrap: 1000

Outputs

• results/consensus/<sample>.fa - masked consensus per sample
• results/consensus/all_consensus.fasta - combined consensuses
• results/aln/wg_alignment.fasta - alignment used for phylogeny
• results/iqtree/wg.treefile - ML tree
• results/coverage/<sample>.depth.txt - per base depth
• results/mutations/<sample>_ag_sites.tsv - amino acid changes in Ag1, Ag2, Ag3
• Optional de novo outputs under results/spades/<sample>/ with QC tables and BLAST hits

Apptainer/Singularity container

HPC-friendly Apptainer/Singularity support is available at containers/Apptainer.def. Build it from the repository root:

apptainer build containers/polio-capsid-ngs-analysis.sif containers/Apptainer.def

Use the image on systems where Apptainer or Singularity is preferred over Docker.

Data governance

See DATA_GOVERNANCE.md for public-data, restricted-data, and sample-identifier handling rules.

How to cite

• Software: Haider SA. Polio Whole-Capsid NGS Surveillance Workflow. Version 1.0.3. Zenodo. https://doi.org/10.5281/zenodo.20257888
• All-version software DOI: https://doi.org/10.5281/zenodo.20257522

References

• Chen S, Zhou Y, Chen Y, et al. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018.
• Li H. Aligning sequence reads with BWA-MEM. 2013. arXiv:1303.3997.
• Danecek P, et al. Twelve years of SAMtools and BCFtools. GigaScience. 2021.
• Garrison E, Marth G. Haplotype-based variant detection from short-read sequencing. arXiv:1207.3907. FreeBayes.
• Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010.
• Picard Toolkit. Broad Institute. 2019.
• Au CH, Ho DN, Kwong A, et al. BAMClipper. Sci Rep. 2017.
• Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7. Mol Biol Evol. 2013.
• Minh BQ, et al. IQ-TREE 2. Mol Biol Evol. 2020.
• Bankevich A, et al. SPAdes. J Comput Biol. 2012.
• Cock PJ, et al. Biopython. Bioinformatics. 2009.
• Kans J. Entrez Programming Utilities Help. NCBI.
• Köster J, Rahmann S. Snakemake. Bioinformatics. 2012.
• Minor PD, Ferguson M, Evans DM, et al. Antigenic structure of polioviruses. J Gen Virol. 1986.
• Hogle J, Chow M, Filman D. Three-dimensional structure of poliovirus at 2.9 Å resolution. Science. 1985.

Contributing

See CONTRIBUTING.md. Open an issue for questions. Do not commit restricted data.

License

MIT. See LICENSE for details.