Installation

Installation

pycoqc-ng is a pure Python package and needs Python 3.9 or newer. It is tested on Python 3.9 through 3.13, on Linux and macOS.

conda (recommended)

A conda environment is the easiest route, because it also brings in pysam and pod5 without needing a compiler:

conda create -n pycoqc-ng -c conda-forge -c bioconda pycoqc-ng
conda activate pycoqc-ng

The bioconda recipe is still in review. Until it is merged, use pip.

pip

pip install pycoqc-ng

To work inside a virtual environment:

python -m venv venv
source venv/bin/activate
pip install pycoqc-ng

From source

git clone https://github.com/duceppemo/pycoqc-ng.git
cd pycoqc-ng
pip install .

For development, install the extras and run the tests:

pip install -e ".[dev,notebook]"
pytest

See Development for more.

Checking the installation

pycoQC --version

should print pycoqc-ng v3.1.0. The package installs six commands:

Command	Purpose
pycoQC	build the QC report
Fastq_to_seq_summary	summary from FASTQ
Bam_to_seq_summary	summary from unaligned BAM (Dorado)
Pod5_to_seq_summary	summary from POD5
Fast5_to_seq_summary	summary from legacy FAST5
Barcode_split	split a summary into one file per barcode

Each takes --help. The package can also be run as a module, python -m pycoQC.

Dependencies

Installed automatically:

numpy ≥1.24, scipy ≥1.10, pandas ≥2.0, plotly ≥5.16, jinja2 ≥3.1, h5py ≥3.10, tqdm ≥4.66, pysam ≥0.22, python-dateutil ≥2.8, pod5 ≥0.2.4

The notebook extra adds ipython, needed only for the jhelp() helper in notebooks. The pdf extra adds kaleido and pypdf, needed only for --pdf_outfile; kaleido drives a headless browser, which is why it is not installed by default:

pip install 'pycoqc-ng[pdf]'

There is no compiled extension. Earlier versions of pycoQC shipped a Cython module; it was replaced by a vectorised NumPy implementation that is both faster and exact, so no build toolchain is required.