Skip to content

Latest commit

 

History

History
56 lines (33 loc) · 1.76 KB

notes_day2.md

File metadata and controls

56 lines (33 loc) · 1.76 KB
Raw

Nextflow course

Day 2

Channels

  • Part between a process and another
  • First-in first-out principle: not necessarily sequential
  • Some steps can start in parallel, maybe the last one in order of the script will start first

Today we will use this github repo: https://github.com/CenterForMedicalGeneticsGhent/nextflow-beginner-feb-2024

First script we will look at is chapter04_channels/01_queue.nf This is a queue channel (.of): we checked the order of the output.

02_value.nf shows a value channel (.value).

04_channel_factory_value.nf shows channel factories.

The only command that creates a value channel is .value, all the other ones create a queue channel (.of, .fromPath, .fromList, etc. )

Processes

Again, execution of a processs has no order.

Directives are settings of a process.

You can also use booleans in a process.

If you do not inlcude a shebang line it will be interpreted automatically as a BASH script.

Working directory

Here you can find:

  • the source code of everything nextflow does behind the scenes. You only need this in case you want to debug a very obscure error
  • command.out : shows the output of the STDOUT
  • all files meaning input files (in the form of a symlink usually) and output files, if any. The symlinks are important because... ?

Staging means what happens to the input files. You can configure this as a directory. By default staging is symlinking the input files.

Operators

They are used to manipulate channels.

If you use it you do not have to use the arrows (->) for the .map function. Because that is the built in

Groovy

Check scripts.

Modularization

You can import any existing module from another file. Modules: collection of processes that you can reuse.