Find the genes that changed, the week the run finishes.

Bulk RNA-seq from FASTQ to a list of differentially expressed genes and the plots to show it, without writing the pipeline yourself.

Join the waitlist

rnaseq.flow5 steps

What is RNA-seq analysis?

RNA-seq analysis turns sequencing reads into a measure of how much each gene is expressed, then tests which genes differ between conditions. A standard bulk pipeline trims the reads, counts them per gene, and fits a statistical model across replicates that reports a fold change and an adjusted p-value for every gene.

Bulk RNA-seq measures the average expression of every gene across all the cells in a sample. It answers comparative questions: which genes respond to a treatment, change over a time course, or differ between a mutant and its wild type. It does not say which cell expressed what; for that, use single-cell RNA-seq.

The analysis has three parts. Quality control removes adapter sequence and poor-quality bases and flags samples that went wrong at the bench. Quantification assigns reads to genes, either by aligning them to the genome with STAR or by estimating transcript abundance directly with Salmon, which is faster and needs far less memory. Differential expression testing, here with DESeq2, models the counts across replicates and reports which genes changed, by how much, and how confident that call is after correcting for testing every gene at once.

Most of the judgement is in the experimental design, not the software. Biological replicates are what let the model separate a real change from ordinary variation between samples: Schurch and colleagues found that three replicates per condition mainly detect the larger changes, and recommend at least six where finding most of the changed genes matters. Anything that differs between samples apart from the condition, such as the day of extraction or the sequencing run, belongs in the design formula, or it will be read as biology.

What happens to your data

For bench scientists comparing conditions, timepoints or genotypes.

Questions it answers

Which genes respond to the treatment?
Do my replicates agree with each other?
Is that knockout doing what I think it is?

1
fastp
Trims adapters and low-quality bases from every read.
Look at: The quality report: a sharp drop along the read, or a high duplication rate.
2
STAR or Salmon
Aligns reads to the genome, or estimates transcript abundance directly.
Look at: The mapping rate. Well below 70% usually means the wrong reference.
3
DESeq2
Tests each gene for a difference between your conditions, corrected for the number of tests.
Look at: The design. Tell it about batches, or a batch will be read as biology.
4
Figures
Draws the PCA, the volcano plot and a heatmap of the top genes.
Look at: The PCA first. Replicates should sit together.

What you take to lab meeting

Follow the protocol step by step

A table of every gene with fold change and adjusted p-value
PCA, volcano and heatmap figures, exportable as SVG
The QC report for every sample

Common questions

How many biological replicates do I need for RNA-seq?

At least three per condition, and more if small changes matter. Schurch et al. (2016) recommend at least six where identifying most differentially expressed genes is the goal. Technical replicates of the same library do not count: the variation that matters is between independent samples.

Should I use STAR or Salmon?

For gene-level differential expression both work well. Salmon estimates transcript abundance without a full alignment, is faster and runs comfortably on a laptop. STAR aligns reads to the genome, which you need if you want to look at reads in a genome browser, find new splice junctions or call variants from RNA.

What does an adjusted p-value mean in DESeq2?

Testing tens of thousands of genes at once guarantees some small p-values by chance. DESeq2 adjusts them with the Benjamini–Hochberg method, so calling genes significant at an adjusted p-value of 0.05 means about 5% of those calls are expected to be false discoveries.

What is a good mapping rate?

For a well-annotated organism, most samples map above about 70%. One sample well below the others usually points to contamination, degraded RNA or a library problem; every sample being low usually means the wrong reference or annotation.

Can I analyse RNA-seq data without a bioinformatician?

Yes, for standard designs. The tools are established and well documented; what takes training is wiring them together in code. On OpenBionode the pipeline is already connected and each setting is explained where you change it, and a bioinformatician can still review every step.

Sources

The papers behind the tools in this pipeline, and the advice on this page. Checked 25 October 2026.

Conesa A, et al. (2016). A survey of best practices for RNA-seq data analysis. Genome Biology 17, 13.
Schurch NJ, et al. (2016). How many biological replicates are needed in an RNA-seq experiment and which differential expression tool should you use? RNA 22, 839–851.
Chen S, Zhou Y, Chen Y, Gu J (2018). fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34, i884–i890.
Dobin A, et al. (2013). STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21.
Patro R, Duggal G, Love MI, Irizarry RA, Kingsford C (2017). Salmon provides fast and bias-aware quantification of transcript expression. Nature Methods 14, 417–419.
Love MI, Huber W, Anders S (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology 15, 550.

Read before you run it

From FASTQ files to a volcano plot: what each step in an RNA-seq pipeline does, what to check after it, and which settings are worth touching.

Your first differential expression analysis, step by step

Explainers, 14 October 2026

Other starter pipelines

Run RNA-seq on your own data.

OpenBionode is in early access. Tell us what you analyse and we’ll make sure the steps you need are ready when your seat opens.