QasimHussain/ppanggolin_pangenomics_helicobacter_pylori

 

Graph-Based Pangenomic Architecture of Helicobacter pylori Using PPanGGOLiN

Organism: Helicobacter pylori | Cohort: n = 41 complete RefSeq assemblies Approach: Partitioned synteny graph pangenomics | Tool: PPanGGOLiN v2.2.6

Abstract

This repository provides a reproducible, end-to-end computational framework for defining the pangenomic landscape of Helicobacter pylori. Moving beyond traditional binary presence/absence matrices, this pipeline leverages PPanGGOLiN (Partitioned PanGenome Graph Of Linked Neighbors) to construct a partitioned synteny graph. This approach unifies gene-family classification with linear genomic neighborhood data. The underlying probabilistic algorithm categorizes the pangenome into three biologically relevant partitions: the Persistent genome (highly conserved structural core), the Shell genome (accessory pathways), and the Cloud genome (rare or strain-specific elements).

The workflow facilitates the identification of Regions of Genomic Plasticity (RGPs) and integration hotspots, establishes a core-genome maximum-likelihood phylogeny, and characterizes the strain-specific resistome using AMRFinderPlus. The analytical pipeline is strictly structured to enhance reproducibility, avoiding intermediate data clutter, and providing publication-ready output.

Biological Context

Helicobacter pylori colonizes the gastric mucosa of a vast segment of the human population, representing a predominant agent of gastric localized diseases and malignancies. A hallmark of its genome is remarkable plasticity; frequent natural transformation, homologous recombination, and horizontal gene transfer drive extensive intra-species diversity.

Conventional analytical methods reliant on simple gene matrices often obscure the structural syntenies underpinning this genomic fluidity. By embedding gene-family variations within an edge-weighted synteny graph, PPanGGOLiN effectively maps out accessory genome dynamics, offering a spatially aware perspective necessary for deciphering H. pylori genomic evolution.

Analytical Pipeline

The streamlined pipeline is structured as follows, emphasizing essential output and discarding non-critical intermediate files:

1. 1.Data Acquisition and Robust Quality Control Acquisition of 41 complete RefSeq genome assemblies using the NCBI Datasets CLI. Stringent all-versus-all Average Nucleotide Identity (ANI) computation via FastANI confirms taxonomic coherence, ensuring all genomes strictly adhere to the >95% threshold for conspecificity.

1. 1.Standardized Genomic Annotation De novo gene calling and functional annotation across all assemblies using Prokka. This ensures uniformity, mitigating biases associated with heterogeneous submission records across institutional sequenced assemblies.

1. 1.Graph Construction and Partitioning Using the ppanggolin panrgp workflow, proteins are clustered, and a directed synteny graph is constructed. A Dirichlet-process mixture model partitions gene families into Persistent, Shell, and Cloud genomes. Regions of Genomic Plasticity (RGPs) are modeled, summarizing recurrent integration events structurally.

1. 1.Core-Genome Phylogeny A core multiple sequence alignment of strictly persistent gene families is generated. A robust maximum-likelihood phylogeny is then inferred via IQ-TREE 2 using optimal substitution models, contextualizing pangenomic variation within the strain evolutionary history.

1. 1.Resistome Profiling Per-strain antimicrobial resistance mapping via AMRFinderPlus, encompassing core resistome determinates, virulence pathways, and stress-response markers.

Repository Structure

The architecture of this repository strictly partitions programmatic logic, input data, and analytical yields:

ppanggolin_pangenomics_helicobacter_pylori/
|
|-- analysis.sh                          # Master shell pipeline executing the complete methodology
|-- visualization.sh                     # Pipeline isolating generation of publication-grade figures
|-- installation.sh                      # Environment setup script for installing bioinformatics dependencies
|-- h_pylori_pangenome_mastery.ipynb     # Interactive cloud-executable optimized Jupyter notebook
|-- README.md                            # Comprehensive project overview and instructions
|
|-- 01_raw_data/                         # Input assembly data and acquisition metadata
|-- 02_qc/                               # Quality control data including FastANI distance matrices
|-- 03_prokka_annotations/               # Primary genomic annotation coordinates and statistics
|-- 04_ppanggolin/                       # PPanGGOLiN graph outputs, partition mapping, and RGP analyses
|-- 05_phylogeny/                        # Persistent family alignments and resultant ML phylogenetic models
|-- 05_resistome/                        # Tabular AMRFinderPlus resistome profiling arrays
|-- 06_visualizations/                   # Executed figures and analytical plot visualizations
`-- bio_env/                             # Local dependency environment structures (large DB files ignored)

Note regarding GitHub publishing: To comply with GitHub file size restrictions, exceptionally large files (>100 MB) corresponding to internal dependencies within the bio_env/ directory have been systematically excluded via the .ignore file configuration. Placeholder README files remain within these directories to document their structural location.

Execution Methodology

The pipeline demands a standard UNIX-like environment (Linux/macOS) equipped with Bash. The environment gracefully handles underlying dependencies via the micromamba or conda ecosystems.

Local Execution

# Clone the repository framework
git clone https://github.com/Qasim-Hussain-Code/ppanggolin_pangenomics_helicobacter_pylori.git
cd ppanggolin_pangenomics_helicobacter_pylori

# Ensure pipeline scripts have execution privileges
chmod +x installation.sh analysis.sh visualization.sh

# Initialize the isolated bio_env environment with all dependencies
./installation.sh

# Activate the environment (use mamba if installed via mamba)
conda activate ./bio_env

# Execute the primary computational pipeline
./analysis.sh

# Synthesize output static figures
./visualization.sh

Cloud Infrastructure ( Environment)

The included .ipynb notebook is architected for direct deployment within . It autonomously manages environment configuration, permitting seamless execution on cloud VMs without specialized prior setups. Execute the cells sequentially from top to bottom on a standard CPU node.

Citation

Utilization of this pipeline structure and embedded methodology should attribute the core algorithmic engine:

Gautreau G, Bazin A, Gachet M, et al. PPanGGOLiN: Depicting microbial diversity via a partitioned pangenome graph. PLOS Computational Biology. 2020;16(3):e1007732. https://doi.org/10.1371/journal.pcbi.1007732