Bulk RNA-seq
Run differential expression on bulk RNA-seq count matrices with guided filtering, DESeq2-style size factors, TMM normalization metadata, VST, sample QC, PCA/sample-distance diagnostics, and TSV exports.
Research question
Which genes are differentially expressed between experimental groups, and what are the effect sizes and significance levels after appropriate normalization?
Who this is for
- Labs running routine bulk RNA-seq DE without maintaining local edgeR/DESeq2 environments
- Cancer center cores processing tumor vs. normal or treatment arm comparisons
- Translational teams needing reproducible bulk pipelines with logged parameters
Data requirements
| Data | Required | Purpose |
|---|---|---|
| Raw or normalized count matrix | Yes | DE input |
| Sample metadata with group labels | Yes | Contrast definition |
| Gene identifiers | Yes | Row names compatible with reference annotations |
| Experimental design / contrasts | Yes | Specify comparison groups |
Bulk analysis is available in the Analyze → Bulk RNA-seq tab and through the analysis API. The guided workflow covers filtering, normalization, differential expression, first-pass sample QC, PCA/sample-distance diagnostics, and exportable result tables.
Workflow
Upload counts + design → Define contrasts → Analyze / Bulk RNA-seq → Review results → Enrichment → Interpret
Step 1 — Prepare count data
Upload a count matrix with samples as columns and genes as rows. Save sample metadata with condition and batch columns on the study Data page. Define contrasts matching your experimental design.
Supported ingestion paths include CSV/TSV count tables and converted pipeline outputs.
Step 2 — Run bulk differential expression
Run the Bulk RNA-seq step from the Analyze page and pick your contrast (e.g. treated_vs_control). Track progress in the Jobs panel.
The bulk pipeline supports DESeq2-style size factors, TMM normalization metadata, VST, gene filtering, differential expression with FDR/log-fold-change filtering, and transformed-expression diagnostics.
Step 3 — Review DE results
Open the completed run's Bulk Results view to examine ranked genes, sample QC, PCA/MDS-style sample mapping, sample-distance heatmaps, and mean-variance diagnostics. Export DE genes, sample QC, distance matrices, and mean-variance tables as TSV files.
Step 4 — Pathway enrichment
Run enrichment on significant gene lists using ORA, or switch to ranked GSEA when you want pathway signal from the full DE table instead of only thresholded genes.
Step 5 — Interpret
Use AI interpretation on completed DE and enrichment runs for plain-language summaries. Review History to inspect run parameters, status, and provenance.
Expected outputs
- Normalized expression estimates
- Differential expression table with statistics per contrast
- Sample QC table with library size, detected genes, and normalization factors
- PCA/sample-distance diagnostics on transformed expression
- TSV exports for DE, sample QC, sample-distance, and mean-variance tables
- Filtered significant gene lists
- GO enrichment results for up/down-regulated genes
- Run record with pipeline version and parameters
Typical analyses
| Analysis | Contrast | Question |
|---|---|---|
| Tumor vs. normal | Primary tumor vs. adjacent normal | What genes are dysregulated in tumor tissue? |
| Treatment response | Post vs. pre treatment | Which pathways shift after drug exposure? |
| Genotype comparison | Mutant vs. wild-type | What is the transcriptional consequence of the alteration? |
| Time course | Timepoint B vs. A | How does expression change over the time series? |
Planned extensions
Dispersion/BCV diagnostics, outlier flagging, set operations, functional-analysis report bundles, publication image exports, and richer cohort design controls remain on the product roadmap. The analysis API remains available for programmatic and core-facility use.