Air Monitoring by Nanopore Sequencing (PRJNA1063692)

🔬 Project Overview

This project utilizes metagenomic analysis of bioaerosols collected via air sampling to monitor microbial communities, employing Oxford Nanopore Technologies (ONT) sequencing. This repository contains a comprehensive workflow designed to process raw sequencing data, perform assembly and binning, and conduct detailed taxonomic and functional analysis.

ENA Project Link: https://www.ebi.ac.uk/ena/browser/view/PRJNA1063692

---

❗️ Important First Step: Data Access & Preprocessing

Before running the main pipeline, you must consult the specific data guide for the dataset you are analyzing. These guides contain critical, non-optional instructions for accessing the correct raw data and preparing your FASTQ files.

• Greenhouse vs. Natural Environment Study (Pilot Study): pilot_study_sample_guide.md
  • Data Format: Raw .fast5 files.
  • Processing: Requires Guppy for basecalling and demultiplexing.
  • Action Required: FASTQ files for Natural Environment samples must be concatenated before use.

guppy_basecaller -i /path/to/input/fast5/ -r -s /path/to/output/ --detect_barcodes -c dna_r10.4.1_e8.2_400bps_hac.cfg

• Urban Air Study: urban_study_sample_guide.md
  • Data Format: Raw .pod5 files.
  • Processing: Requires Dorado for basecalling and demultiplexing.

dorado basecaller dna_r10.4.1_e8.2_400bps_hac@v5.0.0 -r /path/to/input/pod5/ --emit-fastq > basecalled.fastq --kit-name SQK-RBK114-24 --no-trim

The main pipeline assumes you have already completed the steps in the relevant guide and have your demultiplexed FASTQ files ready in a single input directory.

---

📂 Repository Structure

• /bash_scripts: Contains the modular, automated bash pipeline. The main run_pipeline.sh script orchestrates the entire workflow.
• /config: Contains configuration files such as config.yaml with parameters for the pipeline tools.
• /env: Contains the environment.yaml required for creating the conda environment.
• /Python_Scripts: Contains the consolidated Python script (metagenomics_analysis.py) for post-pipeline data analysis and report generation.
• /Taxonomic_and_functional_annotation: Provides detailed guides comparing and explaining the various tools used for taxonomic classification, functional annotation, and AMR gene detection.
• Installation_tutorial.md: A step-by-step guide for installing all required tools and their dependencies.
• pilot_study_guide.md & urban_study_sample_guide.md: Essential guides for initial data access and preparation.

---

🛠️ Installation & Setup

1. Install Tools: For comprehensive installation instructions for all pipeline tools and their dependencies, please refer to the Installation_tutorial.md file. It is highly recommended to use mamba to create a dedicated environment.

2. Download Databases: Before running the pipeline, you must download the necessary databases. A helper script is provided to automate this process.

   # This will download several hundred GB of data
   bash bash_scripts/download_databases.sh

   Important: You must edit the DB_BASE_DIR variable within this script to your desired storage location. After downloading, update the database paths in bash_scripts/run_pipeline.sh.

---

🚀 Usage Workflow

This repository now uses a streamlined, automated pipeline. The manual, step-by-step commands have been replaced by a series of modular scripts orchestrated by a single main script.

1. Configure the Pipeline

Open the main pipeline script, bash_scripts/run_pipeline.sh, and edit the User Configuration section at the top. You must set the correct paths for:

• OUTPUT_BASE_DIR: The main directory where all analysis results will be saved.
• INPUT_FASTQ_DIR: The directory containing your preprocessed, demultiplexed FASTQ files.
• THREADS: The number of CPU threads to use.
• Database Paths: KRAKEN2_DB_PATH, AMRFINDER_DB_PATH, etc. These should match the locations from the download_databases.sh script.

2. Run the Automated Pipeline

Execute the main pipeline script from the root of the repository. It will automatically run all analysis stages in the correct order.

bash bash_scripts/run_pipeline.sh

The pipeline will perform the following stages:

1. Read Processing: Trims adapters, filters reads, generates QC stats, and performs taxonomic classification on reads.
2. Assembly & Polishing: Assembles reads into contigs and polishes them for accuracy.
3. Metagenome Binning: Groups contigs into Metagenome-Assembled Genomes (MAGs).
4. Functional Annotation: Predicts genes and annotates functions for both reads and assemblies, including AMR gene screening.

3. Post-Pipeline Analysis & Reporting

After the main pipeline is complete, use the consolidated Python script to summarize results and generate a final report.

# Example: Generate the final HTML report
python Python_Scripts/metagenomics_analysis.py create_report \
  --input_dir /path/to/your/processing_directory/ \
  --output_file FINAL_REPORT.html

For more details on summarizing specific results (e.g., Kraken2 reports, NanoStat metrics), refer to the README within the Python_Scripts directory.

---

🧰 Tools Integrated

This pipeline integrates the following key bioinformatics tools:

• Basecalling: Guppy / Dorado
• Read Processing: Porechop, NanoFilt
• Assembly: Flye, Minimap2, Racon
• Binning: MetaWRAP (MetaBAT2, MaxBin2, CONCOCT)
• Quality Control: CheckM, NanoStat, Assembly-stats
• Taxonomic Classification: Kraken 2
• Annotation: Prodigal, Prokka, Bakta, eggNOG-mapper
• AMR/Virulence: ABRicate, AMRFinderPlus
• Utilities: Seqkit