votuderep 0.4.1

A CLI tool for dereplicating and filtering viral contigs

votuderep

A Python CLI tool for dereplicating and filtering viral contigs (vOTUs - viral Operational Taxonomic Units) using the CheckV method.

Features

• Dereplicate vOTUs: Remove redundant viral sequences using BLAST-based ANI clustering
• Filter by CheckV metrics: Filter viral contigs based on quality, completeness, and other metrics
• Tabulate reads: Generate CSV tables from paired-end sequencing read directories
• Download training data: Fetch viral assembly datasets for training purposes

Requirements

• Python >= 3.10
• BLAST+ toolkit (specifically blastn and makeblastdb)

Installation

From source

# Clone the repository
git clone https://github.com/yourusername/votuderep.git
cd votuderep

# Install in development mode
pip install -e .

# Or install normally
pip install .

Installing BLAST+

votuderep requires BLAST+ to be installed and available in your PATH:

# Using conda (recommended)
conda install -c bioconda blast

# On Ubuntu/Debian
sudo apt-get install ncbi-blast+

# On macOS
brew install blast

Usage

votuderep provides four main commands: derep, filter, tabulate, and trainingdata.

Dereplicate vOTUs

Remove redundant sequences using BLAST and ANI clustering:

votuderep derep -i input.fasta -o dereplicated.fasta

Options:

• -i, --input: Input FASTA file [required]
• -o, --output: Output FASTA file [default: dereplicated_vOTUs.fasta]
• -t, --threads: Number of threads for BLAST [default: 2]
• --tmp: Temporary directory [default: $TEMP or /tmp]
• --min-ani: Minimum ANI threshold (0-100) [default: 95]
• --min-tcov: Minimum target coverage (0-100) [default: 85]
• --keep: Keep temporary directory with intermediate files

Example:

# Basic dereplication
votuderep derep -i viral_contigs.fasta -o dereplicated.fasta

# With custom parameters
votuderep derep -i viral_contigs.fasta -o dereplicated.fasta \
  --min-ani 97 --min-tcov 90 -t 8

# Keep intermediate files for inspection
votuderep derep -i viral_contigs.fasta -o dereplicated.fasta \
  --keep --tmp ./temp_dir

How it works:

1. Creates a BLAST database from input sequences
2. Performs all-vs-all BLASTN comparison
3. Calculates ANI (Average Nucleotide Identity) and coverage
4. Clusters sequences using greedy centroid-based algorithm
5. Outputs the longest sequence from each cluster (representative)

Filter by CheckV

Filter viral contigs based on CheckV quality metrics:

votuderep filter input.fasta checkv_output.tsv -o filtered.fasta

Required Arguments:

• FASTA: Input FASTA file with viral contigs
• CHECKV_OUT: TSV output file from CheckV

Options:

Length filters:

• -m, --min-len: Minimum contig length [default: 0]
• --max-len: Maximum contig length, 0 = unlimited [default: 0]

Quality filters:

• --min-quality: Minimum quality level: low, medium, or high [default: low]
• --complete: Only keep complete genomes
• --exclude-undetermined: Exclude contigs where quality is "Not-determined"

Metrics filters:

• -c, --min-completeness: Minimum completeness percentage (0-100)
• --max-contam: Maximum contamination percentage (0-100)
• --no-warnings: Only keep contigs with no warnings

Other filters:

• --provirus: Only select proviruses (provirus == "Yes")
• -o, --output: Output FASTA file [default: STDOUT]

Examples:

# Basic filtering - minimum quality
votuderep filter viral_contigs.fasta checkv_output.tsv -o filtered.fasta

# High-quality sequences only
votuderep filter viral_contigs.fasta checkv_output.tsv \
  --min-quality high -o high_quality.fasta

# Complete genomes with minimum length
votuderep filter viral_contigs.fasta checkv_output.tsv \
  --complete --min-len 5000 -o complete_genomes.fasta

# Complex filtering
votuderep filter viral_contigs.fasta checkv_output.tsv \
  --min-quality medium \
  --min-completeness 80 \
  --max-contam 5 \
  --no-warnings \
  --min-len 3000 \
  -o high_confidence.fasta

# Output to stdout (for piping)
votuderep filter viral_contigs.fasta checkv_output.tsv > filtered.fasta

Quality Levels:

CheckV assigns quality levels to viral contigs:

• Complete: Complete genomes (highest quality)
• High-quality: High confidence viral sequences
• Medium-quality: Moderate confidence sequences
• Low-quality: Lower confidence but valid sequences
• Not-determined: Quality could not be determined

The --min-quality option filters inclusively:

• low: Includes Low, Medium, High, and Complete (default)
• medium: Includes Medium, High, and Complete
• high: Includes High and Complete only

Note: "Not-determined" sequences are included by default unless --exclude-undetermined is used.

Tabulate Reads

Generate a CSV table from a directory containing paired-end sequencing reads:

votuderep tabulate reads/ -o samples.csv

Required Arguments:

• INPUT_DIR: Directory containing sequencing read files

Options:

• -o, --output: Output CSV file [default: STDOUT]
• -d, --delimiter: Field separator [default: ,]
• -1, --for-tag: Forward read identifier [default: _R1]
• -2, --rev-tag: Reverse read identifier [default: _R2]
• -s, --strip: Remove string from sample names (can be used multiple times)
• -e, --extension: Only process files with this extension
• -a, --absolute: Use absolute paths in output

Examples:

# Basic usage - generate CSV table
votuderep tabulate reads/ -o samples.csv

# Custom read tags and extension
votuderep tabulate reads/ --for-tag _1 --rev-tag _2 --extension .fq.gz

# Strip patterns from sample names and use absolute paths
votuderep tabulate reads/ --strip "Sample_" --strip ".filtered" -a

Download Training Data

Download viral assembly and sequencing reads for training purposes:

votuderep trainingdata -o ./ebame-virome/

Options:

• -o, --outdir: Output directory [default: ./ebame-virome/]

Example:

# Download to default directory
votuderep trainingdata

# Download to custom directory
votuderep trainingdata -o ./training_data/

Global Options

• -v, --verbose: Enable verbose logging
• --version: Show version and exit
• --help: Show help message

License

MIT License - See LICENSE file for details

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Authors

Andrea Telatin & QIB Core Bioinformatics

©️ Quadram Institute Bioscience 2025