AlexanderLabWHOI/EUKulele

Conda installation issue

susheelbhanu opened this issue · 5 comments

Hi again..

I'm having trouble installing the tool via conda. As noted in one of the earlier issues, I updated conda and still the problems persists.

conda activate eukulele
conda install -c akrinos -c bioconda -c conda-forge EUKulele
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.
Collecting package metadata (repodata.json): done
Solving environment: - Bus error

I tried it on a different high-mem machine and it's been ~11 hours and the environment is still being solved

conda activate eukulele
conda install -c bioconda eukulele
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.
Collecting package metadata (repodata.json): done
Solving environment: |

Any suggestions on how to speed up the process? Thank you!

Hi @susheelbhanu! For the conda installation, I would suggest removing -c akrinos. EUKulele is now available on bioconda, which speeds up the installation! If it still doesn't work on your machine, though, the pip install might be the way to go for you, given your conda setup. Thanks for the queries!

Will give it a go - thank you!

@akrinos Just thought I'd leave this here for future reference. The pip installation worked.

So, I exported the environment file. It's attached here. Hope it helps someone who may run into this. Had to rename it to .txt from .yaml so I could upload. ;)
Updated file here:
eukulele.txt

Thanks for that @susheelbhanu ! But note that if you do use the pip installation, you will need a few other tools installed separately as well, which is the downside of that approach (but it tends to solve faster than conda). Those dependencies are here, and should be sorted with conda install -c bioconda -c conda-forge blast busco=4.0.6 diamond transdecoder via conda (though newer BUSCO should now work and the docs should be updated). Happy to help if anything goes wrong with the dependencies!

Thank you! I suppose one could update the environment file with the dependencies too.. Thanks again for the help!