kristyhoran / coreugate Goto Github PK

It appears there's a debugging println statement that's made it's way into release v2.0.5 which causes errors in Nextflow 23.04.1.
ERROR ~ No such variable: println
It looks like line 97 in coreugate/utils/main.nf

Add to chewie a prep external schema conditional

After a bit of work I have a conda env that works with coreugate. There were a few undocumented dependencies so I thought I'd share my build for others who want to try it.
The main issues were a change in chewbbaca where they removed the --fr argument in v3.0.0, and missing csvtk, rapidnj, cgmlst-dists, and scikit-learn. I'm building the conda env inside a singularity container but this should work for standard conda/mamba as well.

conda create --yes --quiet --name coreugate_2.0.5 -c conda-forge -c anaconda -c bioconda  'nextflow>=20.10' bcftools=1.15.1 singularity 'python>=3.7' 'biopython>=1.70' 'chewbbaca>=2.8.4, <3.0.0' csvtk rapidnj cgmlst-dists scikit-learn

conda activate coreugate_2.0.5

#  use commit 89f613e9984c4d59400ec91d9065a786ddd256dd as there's a println bug in v2.0.5
python3 -mpip install git+https://github.com/kristyhoran/Coreugate.git@89f613e9984c4d59400ec91d9065a786ddd256dd

Clustering is currently complete linkage not single linkage as in readme.

Hi Kristy,
In the old version (https://github.com/MDU-PHL/Coreugate), the input file could be reads or assemblies. However, according to my understanding, in this new version, the input file could only be assemblies. Is this right? Can we input reads files?