Running the nf core rnaseq Example

This page walks through the built-in Demo 2 - nf-core/rnaseq pipeline: the complete nf-core/rnaseq 3.27.0 pipeline, set up on the N-WAVE canvas and preset to the pipeline's small test data.

What the example does

The workflow has three nodes plus notes that explain them:

Node	What it holds
Samplesheet	The pipeline's test samplesheet: 7 sequencing runs of yeast RNA-seq (GSE110004), 4 paired-end and 3 single-end. The reads are downloaded from the nf-core test-datasets repository during the run.
Reference genome (Parameters)	fasta (genome sequence) and gtf (gene annotation) from the nf-core test data.
nf-core/rnaseq (Pipeline)	nf-core/rnaseq 3.27.0 with Use the pipeline's test profile ticked. The samplesheet is connected to --input, the references to --fasta and --gtf.

The pipeline runs with its default settings: read QC (FastQC), trimming (Trim Galore), alignment with STAR and quantification with Salmon (--aligner star_salmon), followed by a MultiQC report.

N-WAVE launches it as

nextflow run nf-core/rnaseq -r 3.27.0 -profile test,docker -params-file params.json --outdir results

with params.json pointing input at the samplesheet and fasta/gtf at the reference URLs.

Requirements

• The Docker installation of N-WAVE (see Installation). The browser demo on GitHub Pages can open, inspect and export the example, but not run it.
• Internet access: Nextflow downloads the pipeline from GitHub, its containers (several GB on the first run) and the test data.
• 4 CPUs and at least 6 GB of memory for Docker. The example's execution settings limit the run to 4 CPUs and 8 GB; N-WAVE lowers these to what the server allows (NWAVE_MAX_CPUS / NWAVE_MAX_MEMORY, default: all CPUs and 80% of the memory). On Docker Desktop, check Settings → Resources.
• Time: about 15–30 minutes on a laptop, plus the first downloads.

Running it

1. Open N-WAVE and click the Demo 2 - nf-core/rnaseq pipeline card on the home page.
2. Double-click the nf-core/rnaseq node to see its settings: the parameters from the pipeline's schema, the launch command and the params file.
3. Click Run Workflow in the bottom bar. The output streams in while the pipeline runs; you can cancel it at any time.
4. When the run finishes, double-click the nf-core/rnaseq node again. Its Results section links the key reports: open multiqc/star_salmon/multiqc_report.html for the summary of every sample (QC, alignment rates, counts). The run panel's Performance tab opens Nextflow's Execution report and Timeline, and shows how long each step took and how much memory it used. All result files lists everything else, e.g. the gene counts in star_salmon/salmon.merged.gene_counts.tsv.

The example is read-only. Changing anything (moving a node, editing a setting, saving) makes an editable copy first; the results of a run on the read-only example are shown until you leave the page, so duplicate it first if you want to keep them with the workflow.

Using your own data

1. Duplicate the example (or change something, which duplicates it).
2. In the Samplesheet node, replace the rows with your runs. Columns: sample (runs with the same name are merged), fastq_1, fastq_2 (empty for single-end reads) and strandedness (auto, forward, reverse or unstranded). Read paths can be URLs, absolute paths on the server, or the names of files uploaded to a File Input node.
3. In the Reference genome node, set fasta and gtf to your organism's files (for example from Ensembl or GENCODE).
4. In the nf-core/rnaseq node, untick Use the pipeline's test profile: it also sets test-only parameters (a test Salmon index, BBSplit references, a Kraken database, ...).
5. Raise Maximum CPU Cores and Maximum Memory in Execution Settings → Resources. A human or mouse genome needs about 40 GB of memory for the STAR index.

Running it outside N-WAVE

Export Project in the bottom bar downloads a folder with run.sh, params.json, inputs/samplesheet.csv and a README. With Nextflow and Docker installed, run bash run.sh in that folder; results go to results/.

Troubleshooting

• A task ran out of memory / requirement exceeds available memory: raise the memory limit in the execution settings, or NWAVE_MAX_MEMORY on the server, and Docker's memory.
• The run stops after 24 hours: that's the default time limit; change Execution Timeout in the execution settings or NWAVE_EXECUTION_TIMEOUT on the server.
• Downloads fail: the pipeline needs access to github.com, raw.githubusercontent.com and the container registries (quay.io, community.wave.seqera.io).

More in Troubleshooting. For the same analysis rebuilt step by step from nf-core modules, see RNA-seq rebuilt from nodes.