replikation / What_the_Phage

WtP: Phage identification via nextflow and docker or singularity
https://mult1fractal.github.io/wtp-documentation/
GNU General Public License v3.0
103 stars 15 forks source link

Ability to enable 'virome' mode for tools: VirSorter + VIBRANT (and others?) #81

Closed mshamash closed 4 years ago

mshamash commented 4 years ago

Hello,

Thank you for making this Nextflow pipeline, it's really a great resource to have!

Something I noticed when doing some preliminary work with WtP so far is that there is no user-facing option to enable the 'virome' modes associated with some tools. For example, VirSorter calls it 'virome decontaination' mode and the flag is --virome, whereas VIBRANT calls it virome mode and the flag is -virome.

Without these flags, these tools assume that we are using a native metagenome as input (which may be the case sometimes, but not all the time for all users) and this could severely affect tool sensitivity and output.

I'm not sure if any other of the included tools have a similar flag, as I am most familiar with VirSorter and VIBRANT. If they do, it may be worth adding as well. Ideally, one Nextflow flag such as --virome would flip "ON" virome mode for all these different tools.

An additional point for VirSorter, it may be useful to have a flag to specify which database to use, since they have both Viromes and Refseq databases included. By using -db 1 which WtP does now, it defaults to the RefSeq database. -db 2 flag is for Viromes and contains RefSeq and more viral sequences on top of that.

Cheers.

Michael

replikation commented 4 years ago
replikation commented 4 years ago

@mult1fractal I'll add this to the v0.8 as its either way part of the parameter optimizations

mshamash commented 4 years ago
  • ah thank you very much for bringing this to our attention
  • a --virome flag seems like the best choice, we take a look how to implement this

    • same for the database

Thanks! On another note, I’ll take this opportunity to ask what the plans are for prophage prediction? Many of these tools also output prophage predictions too. It’d be great to one day be able to input a bacterial genome and get all the predicted prophages by all the tools as output.

replikation commented 4 years ago

yep we are currently looking into this, WtP predicts already prophages - as some tool report them of course (which are then also labelled by the checkV table results as such)

Edit: also some tools "die a bit" if you drop a whole genome in it :)

hoelzer commented 4 years ago

Indeed, good and important suggestions, @mshamash !

@replikation @mult1fractal you might want to have a look here bc/ we also implemented --virome mode for VirSorter in VIRify: https://github.com/EBI-Metagenomics/emg-viral-pipeline/blob/master/nextflow/modules/virsorter.nf

But yeah, it's really just a switch adding --virome or not.

and we are using-db 2 as a default but I am not sure what actually is the best default for most data sets. However, making this an adjustable user input for experienced runs makes totally sense.