The problem. Glucocorticoids are frontline asthma drugs, but which genes they act on in airway smooth muscle — and how — wasn’t fully mapped. Himes and colleagues used RNA-seq to profile treated versus untreated cells and find the responsive genes, landing on CRISPLD2 as a mediator.

The idea. A clean, small, paired design: four cell lines, treated and control, RNA-seq, standard differential expression. Nothing methodologically novel — which is exactly why it became a teaching fixture. The processed counts ship as the Bioconductor airway dataset, the default example for DESeq2 and countless RNA-seq tutorials.

Why it matters. I’ve almost certainly run DESeq2 on these counts while learning the tool, without registering the underlying study. Reading the paper puts biology back under the example: the numbers in the tutorial are a real asthma-drug experiment, not abstract test data. It’s a small correction to how I think about “toy” datasets — they’re someone’s actual result.

Verdict. Modest as biology, outsized as infrastructure — its second life as airway is why it’s on this list. Read it to connect the workflow I practice on to the question it was answering, and to respect that benchmark datasets carry real experimental design worth understanding.