isiteuk 0.0.2

In-situ identification of eukaryotes in metagenomic data

isiteuk

isiteuk classifies genomes by domain (Bacteria, Archaea, Eukaryota) using SingleM marker genes.

Installation

Conda (recommended)

conda create -c conda-forge -c bioconda --override-channels --name isiteuk isiteuk
conda activate isiteuk

PyPI

pip install isiteuk

Non-Python dependencies (SingleM, DIAMOND) must be installed separately — see pixi.toml for the full list.

GitHub (development)

git clone https://github.com/wwood/isiteuk
cd isiteuk
pixi shell
isiteuk --help

This installs isiteuk and all dependencies (including SingleM) into a managed environment via pixi.

Reference data

Download the isiteuk metapackage from Zenodo:

isiteuk data --output-directory /path/to/isiteuk-data
export ISITEUK_METAPACKAGE_PATH=/path/to/isiteuk-data/isiteuk-backpack-0.0.1

Add the export line to your .bashrc or equivalent to avoid repeating it. To verify the download:

isiteuk data --verify-only

Usage

Classify genomes

isiteuk process \
    --output results.tsv \
    --genome-list genomes.txt \
    --threads 64

Or pass genomes directly:

isiteuk process \
    --output results.tsv \
    --genomes genome1.fna.gz genome2.fna.gz \
    --threads 64

The --metapackage flag can override ISITEUK_METAPACKAGE_PATH for a one-off run:

isiteuk process \
    --metapackage /path/to/isiteuk-0.0.1.smpkg \
    --output results.tsv \
    --genomes genome1.fna.gz

To resume an interrupted run:

isiteuk process --continue \
    --output results.tsv \
    --genome-list genomes.txt

Output format

Tab-separated file with one row per genome per domain detected:

genome	domain	num_in_target_domain	num_not_in_target_domain
genome1	d__Bacteria	14.2	0.0
genome2	d__Eukaryota	8.7	1.3

Running tests

Quick tests (no metapackage required):

pixi run test

Full integration tests (requires ISITEUK_METAPACKAGE_PATH to be set):

pixi run test-expensive

Run a single test by name:

pixi run run-a-test test_bacterial_genome