Trying out EUKulele

A minimal working example for EUKulele can be tested using the instructions on this page.

You can obtain sample metagenomic data with simulated MAGs from the following Dropbox link, using the following commands on a Linux system:

wget https://www.dropbox.com/s/l4kvbpqftdad5ib/sample_eukulele.tar.gz?dl=1
tar xzf sample_eukulele.tar.gz?dl=1

Then cd into sample_EUKulele. We have noticed a problem with tarring across systems and through Dropbox. So you may be required to run:

rm samples_MAGs/._sample_*

Now, you should have a clean folder containing only metagenomic samples that you can use for test driving EUKulele!

First, let’s create an environment to run EUKulele in. Create a conda environment and activate it, using:

conda env create -f EUKulele-env.yaml
conda activate EUKulele

This must also be done inside the directory you installed via Dropbox, which contains a conda configuration file that will make sure that every dependency of EUKulele that cannot be installed via pip is available on your system.

You should now download EUKulele via pip using:

python3 -m pip install --index-url https://test.pypi.org/simple/ --no-deps EUKulele

If any dependency is not satisfied, you can install it manually using pip install <requirement> --user. If you install EUKulele via conda instead of via pip, all of the dependencies are installed for you automatically.

Once everything is setup and all dependencies satisfied, from within the sample_EUKulele directory, run:

EUKulele --config curr_config.yaml

Where curr_config.yaml is a configuration file in the directory you downloaded which contains all of the flags needed for a basic EUKulele run. Feel free to open this file if you’re curious, and compare it to the full list of parameters available to customize EUKulele.

Then check the folder test_out_23July in the current directory.