Repository navigation
Nodes
Every node from the + (Add node) menu, by kind. Double-click a node to open its panel: Settings, Code (the Nextflow it adds) and Results (what its tasks produced in the last run). Every node also has a Label.
Kinds: Inputs · Operators · nf-core steps · Custom steps · Output · Note
Files from your computer. Click Upload File(s) on the node, or drop files into its panel,
which also lists, selects and removes them. A file with the same name replaces the old one.
Files up to 256 KB are saved with the workflow; larger ones must be uploaded again after a
reload. Binary files such as .fastq.gz arrive unchanged.
An nf-core-style samplesheet (sample,fastq_1,fastq_2,…), typed in, loaded from a .csv
or started from the rnaseq template. Each row becomes one sample, as nf-core modules
expect it: [ meta, [ reads ] ]. Choose which columns hold the sample id and the reads; the
panel lists the parsed samples and any problems. Reads are names of files uploaded to a File
Input, URLs, or (where Nextflow runs on the computer itself) absolute paths.
Named values (Text, Number, True/false) and Reference files such as a genome
FASTA, GTF or index. Each becomes params.<name>, so it can be changed when the exported
project runs (--fasta …), and node settings can use it as ${params.name}. Every reference
file is an output to connect; its value is an uploaded file name, a path or a URL.
Keeps the lines of text files that match Filter Text: Contains, Starts With, Ends With or Matches (Regex); upper and lower case count as different. Negate keeps the other lines. File Selection limits it to some of the incoming files (none ticked: all).
Changes text: Change Case (upper or lower case) or Replace Text (Find → Replace With).
Joins the files it receives into one, with a Separator between them. Its input takes any number of connections.
Nextflow channel code between nodes, for what plain connections can't do: join a BAM with its
index by sample, mix reports for MultiQC, branch, combine, group or collect. Start from a
Template and edit the Channel code; it reads input.<port> and assigns
output.<port>. The ports are set in Inputs and Outputs.
Installed from the nf-core library.
One tool, such as FastQC, Trim Galore or STAR. Its ports are the module's inputs and outputs, explained under About this nf-core module with links to its documentation. Settings:
- Module inputs: the values the module takes.
-
Extra arguments: options passed to the tool (
ext.args). -
Output prefix: the output file names;
${meta.id}is the sample. -
Save outputs: copies the outputs to
results/<node name>/(on by default). - Resources: CPUs, memory and time for this step (Override defaults).
Several modules wired together by nf-core, such as bam_sort_stats_samtools. Its settings are
the subworkflow's values, what an unconnected input gets, Process config for the processes
inside, and Save outputs.
A complete nf-core pipeline, such as nf-core/rnaseq, run unchanged. The workflow then runs that pipeline instead of a generated script, so it holds only the Pipeline node and the inputs feeding it (Samplesheet, Parameters, File Input). Its settings come from the pipeline's parameter schema; Use the pipeline's test profile runs it on its own test data. See Running the nf-core/rnaseq example.
Your own processes, made in the Custom process editor or with Convert to custom node on any node's Code tab. Their settings are the ones you declared. See Custom nodes and Node packs.
Shows what reaches it. After a run, Results of the last run shows the real files; choose one under View File or all of them, and download it as text, CSV or JSON.
Text on the canvas that explains the workflow. Notes have no ports and add no code.