Seeker integration

[papanikos]

Hi team,

I saw seeker was on your wishlist #1 #75 , it came up on my and figured I 'd give it a go.

A short summary of this PR:

• Dockerfile for seeker
  • I am using a conda env within docker. I was testing this and it seems to work
  • I am not sure how you want to handle containers. It would be nice to have them under one "organization". For now I put in my own Docker hub. Feel free to grab it, rewrite it, whatever.
• Seeker run module, filtering based on author's 0.5 threshold, seeker_wf within main phage.nf
• Inclusion in README, minor stylistic changes in there.
• Addition of reference in the the Citations.txt
• Removal of Docs folder, in favor of docs (maybe a typo in some PR?)

Profiles tested:

• -profile smalltest,local,docker
  • --anno, turn off all tools except deepvirfinder and seeker
• -profile smalltest,local,singularity
  • --anno, turn off all tools except deepvirfinder and seeker
• -profile smalltest,slurm,singularity (our hpc doesn't support docker)
  • full run with all tools, identification and annotation on

All seem to run as expected for me.

Notes

• I haven't really tested performance of seeker on a big dataset. It provides a minimal command line interface without any options for controlling runtime parameters (e.g. cpu usage). Maybe a wrapper as described here included as

run_seeker.py -i <fasta> -o <output.tsv> -c <cores

in the container would add some flexibility. With that said, in the containers config I used parameters similar to other tools that seemed sensible.

Glad to pick any issue during review.

Cheers,
Nikos

[replikation]

• thank you very much for this :) , we will review all the steps, test on big datasets and implement this - that's quite helpful
• you installed docker via conda? can you share the code for this, we currently are writing a "read the docs" and could add this part for conda users :)
• good catch on the --anno part we might have a bug here on our end

[papanikos]

Glad to contribute!

What I meant with "conda env within docker", should be "install conda inside docker (available from the base image anyway) and create an environment inside the docker container"

Some references

• https://pythonspeed.com/articles/activate-conda-dockerfile/
• https://stackoverflow.com/questions/55123637/activate-conda-environment-in-docker

FWIW, I didn't bother with testing if there is any working package available through conda. I had to go through setting up docker on my own Ubuntu 18 laptop.

Singularity and nextflow work though.

$ conda create -n wtp nextflow==20.07.01 singularity==3.6
$ conda activate wtp
(wtp) $ nextflow run replikation/What_the_Phage -r 0.9.0 --setup ...

is my current setup, and it works with -profile local,singularity.

[replikation]

• ah yes i understand now. yep conda install on singularity works, there might be some issues on a cluster environment however compared to a native install of singularity
  thanks

[replikation]

• @mult1fractal looks good on my end