sarek icon indicating copy to clipboard operation
sarek copied to clipboard
verified publisher icon nf-core

TMB

Open FriederikeHanssen opened this issue 3 years ago • 8 comments

#495

Coming up with a POC for the tmb support:

Idea is to keep all relevant config files for the tmb tool in our own assests folder (or any other that we can control for now). Otherwise we are dependent that all of the tools that we would like to support will get added & released upstream. (Over time of course we could contribute them).

For all vcf files the variantcaller & annotation tool needs to be tracked in the meta map. We are already doing this for the variantcallers anyways. So this should be a small and easy change.

Last but not least the respective file can be retrieved with a groovy function for the proper combination.

Now of course we need to start generating all the db and var config files.

I am not sure yet how we could possibly handle users overwritting the config file. From where I stand at the moment we would need a parameter with which you can pass multiple files with a key...

FriederikeHanssen avatar Jul 04 '22 15:07 FriederikeHanssen

Hi @FriederikeHanssen, I had a look at your code ! that's great. For information, there is a typo in the tmb module

        --dbConfig ${dbConfig} \

should be

        --dbConfig ${dbconfig} \

with small 'c'

nservant avatar Jul 20 '22 17:07 nservant

nf-core lint overall result: Passed :white_check_mark: :warning:

Posted for pipeline commit 44d56a1

+| ✅ 151 tests passed       |+
#| ❔   8 tests were ignored |#
!| ❗   1 tests had warnings |!

:heavy_exclamation_mark: Test warnings:

  • pipeline_todos - TODO string in methods_description_template.yml: #Update the HTML below to your prefered methods description, e.g. add publication citation for this pipeline

:grey_question: Tests ignored:

:white_check_mark: Tests passed:

Run details

  • nf-core/tools version 2.6
  • Run at 2022-11-10 23:19:32

github-actions[bot] avatar Aug 03 '22 19:08 github-actions[bot]

  • [ ] Update module in nf-core/modules
  • [ ] vep.yml
  • [ ] strelka.yml
  • [ ] freebayes.yml
  • [ ] Clear up effGenomeSize computation
  • [ ] Update how we handle intervals when running annotation
  • [ ] input starting from tmb directly also needs tbi, but for annotation this would cause issues with channels....
  • [ ] need to propagate all tbis up from variantcalling
  • [ ] allow input configs?

FriederikeHanssen avatar Aug 11 '22 20:08 FriederikeHanssen

Ok this is starting to take shape. A couple of things that I could use help with:

  1. generating the yml files for VEP, Strelka, & Freebayes.
  2. Which parameters do we need?

1 & 2 maybe @apeltzer @tomgutman have any advice?

  1. @maxulysse this is starting to shape into a new subworkflow after annotation, how do we want to call this step? post-processing is not descriptive. We would probably need to track back metadata that we dropped. Here for example we only want to run this tool on somatic data

FriederikeHanssen avatar Nov 10 '22 23:11 FriederikeHanssen

For 3.) --> maybe somehing like "downstream-metrics"? :-)

apeltzer avatar Nov 14 '22 13:11 apeltzer

For 3.) --> maybe somehing like "downstream-metrics"? :-)

I'd even go just for downstream

maxulysse avatar Nov 14 '22 13:11 maxulysse

Hello,

unfortunately I don't have any expertise in VEP, strelka and freebayes. So I can't really help on that. =/ Sorry

tomgutman avatar Nov 16 '22 13:11 tomgutman

I'll prio the generation of the YAML files soon and will either find some time myself or have someone look into it.

apeltzer avatar Feb 02 '23 10:02 apeltzer