nextflow-io / nf-hack18

Nextflow workshop 2018 -- Training pages -> https://nextflow-io.github.io/nf-hack18/

Nextflow workshop 2018

Project repository for the Nextflow workshop that will take place on 22-23 November 2018 in Barcelona.

Location

Centre for Genomic regulation (CRG), PRBB building (you will need an ID card to enter in the building).

Programme

The event programme is available here.

Abstracts

The presentation abstracts are available at this link.

Training

The material for the beginners training on Thursday 22, morning is available as this link.

Event chat

We will use the following Gitter channel during the workshop activities. Feel free to register and use it for any question about logistic, problems, doubts during the workshop, etc.

Poster

Sponsors

nf-hack18's Issues

[HACK TOPIC] implement the ability to import NF scripts in an existing NF pipeline

Already discussed on this issue of the main repository

the best way to parse YAML configuration file

How to remove the work directory automatically

[HACK TOPIC] Process unit testing

Notes from our discussion about how a new generic unit testing module could work:

Present: @ewels, @fstrozzi, @LukeGoodsell, @micans

Testing would essentially act as a wrapper, with three steps:

1. Set up the required files with the storeDir
   • Needs some kind of nextflow syntax to be able to define these test files
2. Run nextflow with a super lenient caching method (file name only?) {new nf feature} so that it skips all of the upstream steps
   • Super lenient caching will mean that we can have empty files for all upstream steps except the penultimate process
   • Don't have staged cached files for the process that we're interested in. Nextflow will run this process for real.
   • Stop the workflow as soon as this process is done {new nf feature}
     • Squash the output channel
     • Required in case something goes wrong and the output filenames change. If this happens then the cache won't be valid and all downstream steps will run again.
3. Check the results from that process

This tool could then be run, specifying which process should be tested. It could then be run in parallel for each process of interest.

How to handle errors (e.g. component, pipeline and network error) and send error message automatically by email