Analysis (Modes, Outputs, Interpretation)

Cellucid’s Analysis panel helps you turn selections/highlights into interpretable results: summary plots, comparisons between groups, and exportable tables/figures.

These pages are written for mixed audiences:

• Wet lab / non-technical: click-by-click recipes and “what success looks like”.

• Computational users: definitions, assumptions, and interpretation pitfalls.

• Power users: caching/state semantics, edge cases, and debugging playbooks.

Important

Analysis results depend on app state.

At minimum, always confirm:

• which view/snapshot is active,

• what is currently visible (filters),

• what groups you’re comparing (highlight pages / derived pages),

• and whether the required data (e.g., gene expression) is available in your dataset.

Tip

New here? Read The analysis lifecycle, end to end first. It walks one analysis from a mouse click to an exported table, in eight numbered steps with a screenshot at each one, and assumes nothing.

If you’re here because “analysis is empty / missing results”, start with Troubleshooting (analysis).

---

Fast path (pick what you’re trying to answer)

You want to…	Start with	When it’s the right choice	Next page
Get a quick read on a selection/group	Quick	You want a fast “is this different?” summary	Analysis mode: Quick (automatic composition + stats)
Inspect distributions and details	Detailed	You need shape/variance/outliers, not just a single number	Analysis mode: Detailed (variable + plots + statistical tests)
See relationships between genes/fields	Correlation	You’re asking “do these move together?”	Analysis mode: Correlation (X vs Y across pages)
Find markers for many groups at once	Marker genes	You want one-vs-rest markers for every group in a categorical field	Analysis mode: Marker Genes (Genes Panel)
Compare genes between two groups	Differential expression	You have gene expression and two pages to compare (A vs B)	Analysis mode: Differential Expression (DE) (Group A vs Group B)
Check a marker panel across groups	Dot plot	You have genes and want mean expression and % expressing per group	Analysis mode: Dot plot
Score a curated gene program	Gene signature	You have a predefined gene set (pathway/module)	Analysis mode: Gene Signature (Gene Signature Score)
Compare cell-type proportions between conditions	Composition	You have cell types, samples and a condition (disease vs healthy)	Analysis mode: Composition

The names in the second column are the labels the app itself uses, exactly as they appear in the panel. The panel groups its rows by what they ask about: Genes (Marker genes, Differential expression, Dot plot, Gene signature), Cells (Composition) and Explore (Quick, Detailed, Correlation).

---

Related concepts (worth skimming first)

• Defining groups: Highlighting and Selection (Groups, Pages, Tools)

• Restricting what’s visible: Filtering (Visibility, Outliers, and Filter Stacks)

• Loading gene expression (and why it can be missing): Data Loading in the Web App (All Paths)

• Exporting plots and publication figures: Figure Export (Publication-Grade SVG/PNG)

• Saving/restoring analysis state (sessions): Session Saving, Restoring, and Sharing

---

Interface reference

The Analysis panel exposes eight analyses in three groups, each with its own validated inputs and result area. The icon on each row undocks that mode into a floating window; the chevron opens it in place.

---

Recommended reading order

1. The analysis lifecycle, end to end — one analysis from start to finish, in eight screenshotted steps. Start here if you have not used the panel before.

2. Analysis mental model (Pages, Variables, Scope) (what analysis operates on + what is cached)

3. Analysis UI overview (where things live; sidebar vs expanded; copy/restore behaviors)

4. Mode pages (03–08, 12, 13) for the specific analysis you need

5. Exporting analysis results (tables/plots, what is reproducible)

6. Troubleshooting (analysis) (symptom → diagnosis → fix)