toxolib 0.1.21

A tool for metagenomic taxonomic profiling and abundance matrix generation

toxolib (v0.1.21)

A Python package for metagenomic taxonomic profiling and abundance matrix generation.

Installation

Using pip

pip install toxolib

Install directly from GitHub

pip install git+https://github.com/dhruvac29/toxolib.git

Using conda

We recommend using conda to install all dependencies. An environment file is included in the package:

# Clone the repository
git clone https://github.com/dhruvac29/toxolib.git
cd toxolib

# Create and activate the conda environment
conda env create -f environment.yml
conda activate taxonomy_env

# Install the package
pip install -e .

Requirements

This package requires the following external tools to be installed and available in your PATH:

• Kraken2
• Bracken
• Krona (for visualization)
• fastp (for preprocessing)
• bowtie2 (for host removal)
• samtools

All these dependencies are included in the conda environment file.

Database Setup

Automated Database Setup

Toxolib provides automated database setup for both local and HPC environments.

Local Database Setup

# Set up both Kraken2 and corn genome databases
toxolib db-setup -o /path/to/databases --kraken --corn

# Set up only Kraken2 database
toxolib db-setup -o /path/to/databases --kraken

# Set up only corn genome database
toxolib db-setup -o /path/to/databases --corn

# Force re-download of databases even if they exist
toxolib db-setup -o /path/to/databases --kraken --corn --force

After setup, you should set the environment variable for Kraken2:

export KRAKEN2_DB_DIR=/path/to/databases/Kraken2_DB

HPC Database Setup

When submitting jobs to the HPC, you can automatically download and set up the databases locally and upload them to the HPC:

# Automatically download locally and upload both databases to the HPC
toxolib hpc -r sample1_L001_R1.fastq.gz sample1_L001_R2.fastq.gz sample1_L002_R1.fastq.gz sample1_L002_R2.fastq.gz -o /path/on/hpc/output_dir \
    --setup-kraken-db --setup-corn-db

# Automatically download locally and upload only Kraken2 database
toxolib hpc -r sample1_L001_R1.fastq.gz sample1_L001_R2.fastq.gz sample1_L002_R1.fastq.gz sample1_L002_R2.fastq.gz -o /path/on/hpc/output_dir \
    --setup-kraken-db

# Automatically download locally and upload only corn genome database
toxolib hpc -r sample1_L001_R1.fastq.gz sample1_L001_R2.fastq.gz sample1_L002_R1.fastq.gz sample1_L002_R2.fastq.gz -o /path/on/hpc/output_dir \
    --setup-corn-db

When using these options, toxolib will:

1. Download the databases to your local machine
2. Extract the databases locally
3. Upload the extracted databases to the HPC
4. Configure the Snakefile to use the correct database paths

This approach works even if your HPC has restricted internet access or firewalls that prevent direct downloads.

Manual Database Setup

If you prefer to set up the databases manually, you can follow these steps:

Kraken2 Database

You can download the standard Kraken2 database from: https://genome-idx.s3.amazonaws.com/kraken/k2_standard_20240112.tar.gz

wget https://genome-idx.s3.amazonaws.com/kraken/k2_standard_20240112.tar.gz
tar -xzf k2_standard_20240112.tar.gz -C /path/to/kraken2/database
export KRAKEN2_DB_DIR=/path/to/kraken2/database

Corn Genome Database

For host removal, you can download the corn genome reference from: https://glwasoilmetagenome.s3.us-east-1.amazonaws.com/corn_db.zip

wget https://glwasoilmetagenome.s3.us-east-1.amazonaws.com/corn_db.zip
unzip corn_db.zip -d /path/to/corn_db

Usage

Local Usage

Generate abundance matrix from raw data

toxolib abundance -r raw_data_1.fastq.gz raw_data_2.fastq.gz -o output_directory

This will:

1. Run Kraken2 on the raw data
2. Run Bracken on the Kraken2 results
3. Generate an abundance matrix from the Bracken results

Create abundance matrix from existing Bracken files

toxolib matrix -i sample1_species.bracken sample2_species.bracken -o abundance_matrix.csv

HPC Usage

Toxolib can run the analysis pipeline on an HPC cluster using SLURM for job scheduling.

1. Set up HPC connection

toxolib hpc-setup --hostname your-hpc-server.edu --username your-username --key-file ~/.ssh/id_rsa

This will save your HPC connection details to ~/.toxolib/hpc_config.yaml.

2. Run the pipeline on HPC

toxolib hpc -r raw_data_1.fastq.gz raw_data_2.fastq.gz -o /path/on/hpc/output_dir \
    --kraken-db /path/on/hpc/kraken2_db \
    --corn-db /path/on/hpc/corn_db \
    --partition normal --threads 32 --memory 200 --time 144:00:00

This will:

1. Upload your raw data files to the HPC
2. Create a Snakemake workflow file
3. Upload an environment.yml file to the HPC
4. Submit a SLURM job to run the analysis
5. Return a job ID for tracking

Automatic Conda Environment Creation

When submitting a job to the HPC, toxolib will automatically:

1. Upload a conda environment.yml file to the HPC
2. Create a conda environment in the output directory if it doesn't exist
3. Activate the environment before running the analysis

This ensures all required dependencies are available on the HPC without requiring manual environment setup.

3. Check job status

toxolib hpc-status --job-id your_job_id

4. Download results when complete

toxolib hpc-download --job-id your_job_id --output-dir ./local_results

5. HPC File Management

Toxolib provides several commands to manage files and directories on the HPC:

# Print working directory on HPC
toxolib hpc-pwd

# Change directory on HPC
toxolib hpc-cd --path /path/on/hpc

# List files on HPC
toxolib hpc-ls [--path /path/on/hpc] [--long] [--all]

# Create directory on HPC
toxolib hpc-mkdir --path /path/on/hpc [--parents]

These commands allow you to navigate and manage files on the HPC without needing to manually connect via SSH.

6. Persistent Connections

Toxolib now maintains persistent SSH connections to the HPC between commands. This means:

1. You only need to authenticate once per terminal session
2. Connections remain open for subsequent commands
3. No need to reconnect for each command
4. Faster command execution since connection setup is skipped

Connections are automatically managed:

• Kept alive with keep-alive packets
• Automatically closed after 10 minutes of inactivity
• Automatically closed when your terminal session ends

If you prefer not to use persistent connections, you can disable this feature with the --no-persist flag:

toxolib hpc-pwd --no-persist

7. Session-Based Password Storage

Toxolib includes session-based password storage, which allows you to enter your password just once per terminal session. The password is securely stored in memory and automatically reused for subsequent commands during your session.

• Password is only stored for the current terminal session
• Password is automatically cleared when your session ends
• Password is encrypted in memory using strong cryptography
• No passwords are ever saved to disk

This feature makes it more convenient to run multiple HPC commands without having to re-enter your password each time.

Get Current Working Directory

# Show current directory and close connection
toxolib hpc-pwd

# Show current directory and keep connection open
toxolib hpc-pwd --keep-open

Change Directory

# Change directory and close connection
toxolib hpc-cd --path /path/to/directory

# Change directory and keep connection open
toxolib hpc-cd --path /path/to/directory --keep-open

# Go up one level
toxolib hpc-cd --path ..

# Go up one level and keep connection open
toxolib hpc-cd --path .. --keep-open

Create Directory

# Create directory and close connection
toxolib hpc-mkdir --path /path/to/new/directory

# Create directory and keep connection open
toxolib hpc-mkdir --path /path/to/new/directory --keep-open

List Files

# List files in current directory and close connection
toxolib hpc-ls

# List files in current directory and keep connection open
toxolib hpc-ls --keep-open

# List files in specific directory
toxolib hpc-ls --path /path/to/directory

# List files in specific directory and keep connection open
toxolib hpc-ls --path /path/to/directory --keep-open

# Long format listing (like ls -l)
toxolib hpc-ls --long

# Show hidden files (like ls -a)
toxolib hpc-ls --all

# Combine options
toxolib hpc-ls --path /path/to/directory --long --all

# Combine options and keep connection open
toxolib hpc-ls --path /path/to/directory --long --all --keep-open

Manual Setup on HPC

When using the HPC functionality, you can manually upload and extract these databases on your HPC system:

# On your local machine, download the databases
wget https://genome-idx.s3.amazonaws.com/kraken/k2_standard_20240112.tar.gz
wget https://glwasoilmetagenome.s3.us-east-1.amazonaws.com/corn_db.zip

# Upload to HPC (using scp)
scp k2_standard_20240112.tar.gz [e-mail]:/path/on/hpc/
scp corn_db.zip [e-mail]:/path/on/hpc/

# SSH into HPC and extract
ssh [e-mail]
mkdir -p /path/on/hpc/kraken2_db
tar -xzf /path/on/hpc/k2_standard_20240112.tar.gz -C /path/on/hpc/kraken2_db
mkdir -p /path/on/hpc/corn_db
unzip /path/on/hpc/corn_db.zip -d /path/on/hpc/corn_db

Then when running toxolib, specify these paths:

toxolib hpc -r raw_data_1.fastq.gz raw_data_2.fastq.gz -o /path/on/hpc/output_dir \
    --kraken-db /path/on/hpc/kraken2_db \
    --corn-db /path/on/hpc/corn_db

License

MIT