Back to Case Studies

Case study · Bulk RNA-seq

Dexamethasone Response in Human Airway Smooth Muscle Cells

Eight airway smooth-muscle RNA-seq samples were analyzed with a cell-line-aware DESeq2 model. The evidence record separates treatment-associated expression changes from pathway or causal claims that the study cannot establish.

Homo sapiens8 samples4 cell linesDESeq2
View full Pipette session Run a similar analysis

Original 8-page session report · PDF · 860 KB

At a glance

A treatment contrast that accounts for baseline cell-line differences

Four dexamethasone-treated and four control samples span four cell lines. Including cell line in the model avoids treating all eight samples as exchangeable biological replicates.

Samples8
Cell lines4
Genes tested22,369
FDR < 0.054,000
01 / Scientific context

Which genes change after dexamethasone treatment while controlling for cell line?

The input contained 63,677 gene-level count rows. Genes with fewer than 10 total counts were removed before modeling; 22,369 genes remained.

What treatment-associated gene-expression changes are detectable across four human airway smooth-muscle cell lines after accounting for line-specific baselines?
Samples4 treated and 4 control
Cell lines4
ModelDESeq2 design ~ cell + dex
SoftwareDESeq2 1.40.2 in R 4.3.0
FilterAt least 10 total counts across samples
02 / Analysis workflow

Count filtering, normalization, modeling, and diagnostics

Inspect counts and metadataMatched treatment labels and cell-line identifiers across eight samples.
Filter low countsRetained 22,369 genes with at least 10 total counts.
Fit DESeq2Applied median-of-ratios normalization, a cell-plus-treatment design, Wald tests, and BH correction.
Audit separationUsed variance-stabilized counts for PCA and generated volcano, MA, and heatmap diagnostics.
Statistical boundaryThe reported significant set uses adjusted p < 0.05. No minimum effect-size threshold is documented for the 4,000-gene total.
03 / Results

The treatment contrast produced broad expression changes

4,000 significant genes

At adjusted p < 0.05, 2,193 increased and 1,807 decreased.

Stronger thresholds

2,901 passed adjusted p < 0.01 and 2,073 passed adjusted p < 0.001.

PCA separation

PC1 explained 48% and PC2 explained 23% of the displayed variance.

Cell line modeled

The design included cell line directly instead of pooling unmatched replicates.

PCA separating dexamethasone-treated and control airway smooth muscle samples
Figure 1. Variance-stabilized sample PCA. The first two components account for 71% of displayed variance. Separation supports a treatment-associated signal but is not a pathway-level result.

The significant set contains both increases and decreases

Volcano plot of dexamethasone differential expression in airway smooth muscle
Figure 2. Differential-expression volcano plot. The report shows a broad treatment-associated signal across the tested genes.

Top genes cluster samples by treatment in the report

Heatmap of the top differentially expressed genes after dexamethasone treatment
Figure 3. Top-gene expression heatmap. The displayed pattern is descriptive and depends on the selected genes and transformation.
04 / Limitations

What the eight-sample analysis cannot establish

  • The dataset contains only eight samples across four cell lines.
  • No additional covariates or cell-line-by-treatment interaction were modeled.
  • The report does not provide a public accession or independent experimental validation.
  • Pathway enrichment is not documented, so gene-level results should not be presented as a reproduced mechanism.
  • Differential expression is an association with treatment and does not establish causality.
Interpretation boundary

The strongest supported conclusion is a broad dexamethasone-associated expression shift after controlling for cell line. Mechanistic and clinical claims require separate evidence.

05 / Reproducibility and outputs

Artifacts recorded in the Pipette session

de_results.csvComplete DESeq2 results
de_significant.csvFDR-significant genes
normalized_counts.csvNormalized values for 22,369 genes
deseq2_object.rdsFitted analysis object
de_config.jsonConfiguration and thresholds
provenance.jsonExecution provenance
Original Pipette session report8-page PDF · 860 KB · generated February 10, 2026. The HTML article is the primary case study; the PDF is the preserved session record.
View session Download PDF
RNA-seq methods
TPM vs FPKM vs Counts
Pathway analysis
GSEA vs GO Enrichment

More RNA-seq case studies

Rice salt-stress RNA-seq analysis →

Run a similar differential-expression analysis

Bring a count matrix and sample metadata, then describe the contrast and covariates that matter.

Start in Pipette