Analyze
Analyze runs computational workflows. Tasks are grouped by research intent rather than by software package name.
Overview
Shows task readiness (which steps have completed runs), recommended next actions, and links to stale outputs.
When uploaded AnnData files contain Bioconductor-style layers, relevant steps expose a matrix-source control. Counts-oriented workflows default to detected counts, while expression-oriented workflows default to logcounts or normalized matrices when available. The selected source is recorded in run parameters.
Find Structure
| Step | Purpose |
|---|---|
| QC | Filter cells by counts, genes, and mitochondrial percentage |
| Normalization | Log-normalize, scale, select highly variable genes |
| Clustering | Neighbors graph, Leiden clustering, UMAP |
| Spatial domains | Cluster spatial spots into tissue domains, rank spatially variable genes, summarize neighborhood enrichment, compare domains by condition, and nominate co-localization niche candidates |
Steps must generally run in order; each saves a checkpoint .h5ad for the next step.
Compare Groups
| Step | Purpose |
|---|---|
| Differential expression | Wilcoxon rank-sum on contrasts or cluster markers |
| Pseudobulk DE | Aggregate single-cell counts by sample and cell type/cluster, prefill saved contrasts, run replicate-aware per-group DE, and review volcano/table results with TSV exports |
| Pathway enrichment | ORA, ranked GSEA, ssGSEA/GSVA gene-set scoring, run-scoped custom gene sets, and result gene-set comparison from DE results or expression profiles |
For contrast-based DE and pseudobulk DE, save contrasts on the Data page first. Cluster marker mode compares each cluster to all others.
Bulk RNA-seq
| Step | Purpose |
|---|---|
| Bulk RNA-seq | Sample-level count filtering, DESeq2-style/TMM normalization metadata, VST, differential expression, sample QC, PCA/sample-distance diagnostics, and TSV exports |
Bulk results open in a dedicated view with ranked DE genes, sample QC, PCA/MDS-style sample mapping, sample-distance heatmap data, and mean-variance diagnostics.
Build Predictors
| Step | Purpose |
|---|---|
| Biomarker discovery | Feature selection (mRMR or random forest), optional WGCNA, classifier CV |
| WGCNA Modules | Standalone co-expression modules with soft-threshold diagnostics, hub genes, eigengene/trait review, lightweight hub-edge graph, and TSV exports |
| Coverage DEG | Pairwise class comparisons, cross-contrast DEG recurrence ranking, inline disease evidence, and TSV exports |
Requires a class label column in obs (default condition) with at least two classes.
Integrate / Scale
| Step | Purpose |
|---|---|
| Batch Integration | First-pass PCA batch centering with integrated clustering, before/after PCA previews, batch-mixing metrics, biology-preservation checks, and confounding warnings |
| Cell Type Annotation | Assign auditable cluster-level labels from the built-in PBMC/immune marker reference with compatibility checks, score heatmaps, confidence/delta review, and warnings |
| Trajectory & Cell Cycle | Score cell-cycle programs, compute graph-based pseudotime, review UMAP overlays, inspect assumption warnings, and rank genes associated with pseudotime |
| Sketch & Scale | Create representative sketches, projection artifacts, downsampled previews, dedicated sketch results, and large-dataset guardrails |
| Reference Deconvolution | Estimate cell-type fractions from sample-level or spot-level expression |
Result pages
Completed steps link to dedicated result views:
- DE Results — volcano plot and sortable gene table
- Pseudobulk Results — per-cell-type/group selector, volcano plot, sample aggregate table, skipped-group warnings, and TSV exports
- Integration Results — before/after PCA previews, batch-mixing metrics, batch composition, and correction warnings
- Annotation Results — reference compatibility, per-cluster score heatmap, confidence/delta review, and supporting markers
- Trajectory Results — pseudotime/cell-cycle UMAP overlays, assumption warnings, cluster ordering, pseudotime-associated genes, and TSV exports
- Sketch Results — representative-cell preview, projected cluster counts, scale summary, and artifact status
- Bulk Results — sample QC, PCA/sample-distance diagnostics, mean-variance table, and DE gene table
- Biomarker Results — ranked genes and classifier metrics
- Spatial Results — spot/domain viewer, SVG ranked list and overlays, neighborhood enrichment, condition-domain comparison, niche candidates, and TSV exports
- WGCNA Module Explorer — module summaries, hub genes, eigengene/trait tables, selected-module hub-edge graph, and TSV exports
- Coverage DEG Results — recurrent class-comparison genes, pairwise DE details, inline disease evidence, and TSV exports
- Enrichment Results — GO/KEGG, GSEA, and score-summary tabs with top-term bars, per-term gene exports, and pairwise set operations across available result gene lists
DE, pseudobulk, bulk, enrichment, biomarker, and spatial result views include TSV export buttons for supplement-ready tables.
Errors and reruns
Failed runs show the error message on Analyze and in expandable rows on History. Fix parameters or input data, then re-run. Re-running QC or normalization invalidates downstream checkpoints.