Skip to content

README.md

Latest commit

 

History

History
115 lines (82 loc) · 4.5 KB

README.md

File metadata and controls

115 lines (82 loc) · 4.5 KB

Identifying viral OTUs from metagenomic assemblies

This is a bioinformatics pipeline for identifying non-redundant viral populations from metagenomic assemblies, leveraging VirSorter2, CheckV, and geNomad. Predicted viral sequences are filtered, aggregated across samples, and dereplicated at 95% ANI over 85% aligned fraction to generate a set of viral operational taxonomic units (vOTUs) for downstream analyses.

1. Overview

This pipeline performs the following steps:

  1. VirSorter2+CheckV
    • a. Viral prediction (VirSorter2 with options --keep-original-seq)
    • b. Quality assessment and host trimming (CheckV)
    • c. Filtering to retain viral sequences ≥ 5kbp and viral ≥ host genes
  2. geNomad
    • a. Viral prediction (geNomad with options --enable-score-calibration)
    • b. Filtering to retain viral sequences ≥ 5kbp and FDR < 1%
  3. Aggregation: Combines viral sequences across samples.
  4. Dereplication: Clusters viral sequences at 95% ANI over 85% aligned fraction.

2. Directory structure

viral-identification/
├── README.md
├── vir-id.smk
├── config.yaml
├── samples.tsv  # table of sample IDs and metagenomic assembly path names
└── results/     # outputs (not included)
data/
└── contigs/     # metagenomic assemblies (not included)
resources/
└── genomad_db/  # geNomad database (not included)

3. Dependencies

This pipeline is built using Snakemake. All software dependencies are managed via conda.

  • Snakemake (tested with v5.26.0)
  • VirSorter2 (tested with v2.2.4)
  • CheckV (tested with v1.0.1)
  • geNomad (tested with v1.8.1)
  • CD-HIT (tested with v4.8.1)
  • python (tested with v3.10.0)
  • pandas (tested with v2.2.0)
  • Biopython (tested with v1.83)

4. System requirements

This pipeline is designed for high-performance computing (HPC) environments.

Hardware requirements

  • Memory: Minimum 64 GB RAM
  • CPU: Scalable from 16 to 48+ cores

5. Installation & Setup

5.1 Clone the repository

git clone https://github.com/CSB5/SPMP_Phages.git

5.2 Install dependencies

conda create -n vir-id -c conda-forge -c bioconda snakemake=5.26.0 virsorter=2.2.4 checkv=1.0.1 genomad=1.8.1 cd-hit=4.8.1 pandas=2.2.0 biopython=1.83
conda activate vir-id

5.3 Download databases

6. Usage

The pipeline requires a set of metagenomic assembly files as input. Modify sample_id, contigs_path, and contig_header_prefix in samples.tsv accordingly.

The following parameters are defined in config.yaml:

  • Minimum viral sequence length (min_len, default: 5000)
  • VirSorter2 minimum score (min_vs2_score, default: 0.5)
  • geNomad maximum FDR (max_gmd_fdr, default: 0.01)
  • geNomad database path (genomad_db)

Navigate to the project subdirectory to run the pipeline:

cd pipelines/viral-identification
snakemake -s vir-id.smk --cores 48

7. Outputs

All results are saved in the results/ directory. The primary outputs are:

  • viral.fna: Viral sequences identified by VirSorter2+CheckV and geNomad, aggregated across all samples, and dereplicated at 100% ANI.

  • v95.fna: Non-redundant vOTUs at 95% ANI over 85% aligned fraction.

These outputs are used as inputs for downstream viral clustering, host association and other analyses.

8. Citation

If you use this pipeline in your research, please cite:

  • Chen, H. et al. GuFi phages represent the most prevalent viral family-level clusters in the human gut microbiome. bioRxiv. https://doi.org/10.64898/2026.01.26.701711
  • VirSorter2: Guo, J. et al. VirSorter2: a multi-classifier, expert-guided approach to detect diverse DNA and RNA viruses. Microbiome 9, 37 (2021).
  • CheckV: Nayfach, S. et al. CheckV assesses the quality and completeness of metagenome-assembled viral genomes. Nature Biotechnology 39, 578–585 (2021).
  • geNomad: Camargo, A. P. et al. Identification of mobile genetic elements with geNomad. Nature Biotechnology 42, 1303–1312 (2024).
  • CD-HIT: Fu, L. et al. CD-HIT: accelerated for clustering the next-generation sequencing data. Bioinformatics 28, 3150–3152 (2012).