TODO: COMPLETE FastPLong: Quality Control for Long Read Sequencing Data

Setting Up FastPLong with Conda and Docker

FastPLong is an ultra‑fast preprocessing and quality‑control tool for long‑read sequencing data (Nanopore, PacBio, Cyclone, etc.). It generates filtered FASTQ output along with HTML and JSON QC reports.

In this guide, we'll cover:

• Installing FastPLong via Conda
• Downloading the prebuilt binary
• Compiling from source (optional)
• Running FastPLong in a Docker container
• A simple example workflow using a FASTQ file

1. Prerequisites

• Conda (Miniconda or Anaconda) for package installs
• Docker (optional for containerized use)
• A sample long‑read FASTQ file (plain or compressed)

2. Installing FastPLong via Conda

FastPLong is available on Bioconda:

# Ensure Bioconda channel is enabled
conda config --add channels defaults
conda config --add channels bioconda
conda config --add channels conda-forge

# Create and activate env
conda create -n fastplong_env python=3.9 -y
conda activate fastplong_env

# Install FastPLong
conda install fastplong -y

To upgrade later:

conda update fastplong -y

3. Downloading the Prebuilt Binary

For Linux users, you can grab the latest standalone binary:

# Download latest (CentOS build)
wget http://opengene.org/fastplong/fastplong \
     -O fastplong
chmod +x fastplong

# Or fetch a specific version
wget http://opengene.org/fastplong/fastplong.0.2.2 \
     -O fastplong
chmod +x fastplong

You can now run ./fastplong --help to see all options.

4. Compiling from Source (Optional)

If you need the bleeding‑edge version or wish to customize:

# Install dependencies
conda install -c conda-forge libdeflate isa-l libhwy -y

# Clone & build
git clone https://github.com/OpenGene/fastplong.git
cd fastplong
make -j$(nproc)
make test
sudo make install

This installs fastplong into your system PATH.

5. Running FastPLong in Docker

To avoid local installs, use a containerized setup. Create a simple Dockerfile:

# Dockerfile
FROM continuumio/miniconda3:latest

# Install FastPLong
RUN conda config --add channels defaults \
    && conda config --add channels bioconda \
    && conda config --add channels conda-forge \
    && conda install fastplong -y \
    && conda clean --all -y

ENTRYPOINT ["fastplong"]

Build and tag:

docker build -t fastplong:latest .

Run on your FASTQ:

docker run --rm -v $(pwd):/data fastplong:latest \
  -i /data/reads.fastq.gz \
  -o /data/filtered.fq.gz \
  --html /data/report.html \
  --json /data/report.json

6. Example Usage

Basic command

fastplong -i reads.fastq.gz -o filtered.fq.gz

This applies default filters and writes:

• filtered.fq.gz: reads passing all filters
• fastplong.html: interactive HTML QC report
• fastplong.json: machine‑readable JSON report

Common options

• Quality filters (default on):
  • -m 10 (mean quality ≥Q10)
  • -q 15 -u 40 (phred ≥15; ≤40% low‑quality bases)
• Length filters:
  • -l 1000 (min length 1 kb)
  • --length_limit 50000 (max length 50 kb)
• Adapter trimming:
  • -s START_SEQ -e END_SEQ
  • -a adapters.fasta
• Streaming:
  • --stdout to pipe passing reads into another tool
  • --stdin to read from STDIN
• Splitting:
  • --split 4 (into 4 output files)
  • --split_by_lines 100000 (100 k lines per file)

Example with more filters:

fastplong \
  -i reads.fastq.gz \
  -o filtered.fq.gz \
  --failed_out failed.fq.gz \
  -m 12 -l 500 \
  --cut_front --cut_tail \
  --split 8 \
  --html qc.html --json qc.json

7. Tips & Best Practices

• Preview a subset: use --reads_to_process 100000 for a quick QC check.
• Avoid overwriting: add --dont_overwrite to protect existing outputs.
• Parallelization: combine --threads (via environment) with --split for concurrent downstream steps.
• Low‑complexity filtering: enable -y for simple repeats removal.
• Report customization: change titles with -R "My FastPLong Report".

8. Conclusion

FastPLong streamlines long‑read QC and preprocessing, offering flexible filters, streaming, and detailed reports. Whether installed via Conda, binary, source, or Docker, you can integrate it seamlessly into your analysis pipeline.

Happy filtering!