Overview
Computational Biology workflows connect data ingestion, guided pipelines, interactive exploration, interpretation, and reproducible history in a single study workspace. Each use case below maps a common research question to platform capabilities, data requirements, and a suggested analysis path.
Use case index
| Use case | Research question | Key capabilities |
|---|---|---|
| Single-cell atlas | What cell types exist, how do they mature or change, and how do they differ between conditions? | QC, clustering, sketch scaling, integration, annotation, pseudobulk DE, trajectory, enrichment, history |
| Single-cell CNV and malignant cells | Which cells or clusters show expression-derived chromosome-scale CNV signal consistent with malignant tumor populations? | Reference-cell QC, chromosome-ordered CNV heatmap, malignant score, event calling, subclones, state calls, oncology handoff |
| Spatial transcriptomics | Where are genes expressed in tissue and what spatial domains exist? | Spatial domain clustering, neighborhood enrichment, reference deconvolution, spot viewer, gene overlays |
| Biomarker discovery | Which genes best classify samples and how well does a classifier perform? | WGCNA modules, coverage DEG, disease-evidence gene sets, mRMR feature selection, cross-validated classifiers |
| Bulk RNA-seq | Which genes are differentially expressed between groups? | Guided filtering, DESeq2-style size factors, TMM metadata, VST, differential expression, reference deconvolution |
Who these use cases serve
| Audience | Typical goals |
|---|---|
| Wet-lab biologists | Guided analysis without writing R or Python |
| Computational biologists | Parameter transparency, reproducibility, and provenance |
| Bioinformatics cores | Repeatable client-facing workflows at scale |
| Translational and biotech teams | DE, biomarker panels, and reproducible analysis history |
| Pharma computational biology groups | Standardized pipelines with audit trails and recorded parameters |
Common data requirements
Most use cases start with a study containing:
- Datasets —
.h5ad, 10x outputs, count matrices, or bulk expression tables - Metadata — sample and condition columns merged into
obsbefore analysis - Contrasts (for DE and biomarker) — group comparisons defined on the Data page
Choosing a starting point
| If your primary data is… | Start with… |
|---|---|
Single-cell RNA-seq (10x, .h5ad) | Single-cell atlas |
| Tumor single-cell RNA-seq needing malignant labels | Single-cell CNV and malignant cells |
Visium or spatial .h5ad | Spatial transcriptomics |
| Labeled samples for classification | Biomarker discovery |
| Bulk count matrix with group labels | Bulk RNA-seq |
Study workflow sections
All use cases follow the same left-to-right flow:
Data → Explore → Analyze → Interpret → History
See Study workflow for section details.
What these use cases are not
These workflows support exploratory and translational research. AI interpretation summarizes computed outputs — it does not perform genomics statistics or fabricate metrics.
Next steps
- Quick start — run your first single-cell study
- Study workflow — how the six sections fit together
- Glossary — key terms and concepts