Annotation Sources & Instructions

How to Use Gene Annotate

Single Gene: Type a gene symbol (e.g. TP53, BRCA1) into the search box and press Enter or click Search. Switch species between Human and Mouse using the buttons above the search bar. You can also right-click any gene label across OmicsPlot and select Annotate this gene to open it directly here.

Batch Annotation: Switch to the Batch tab, paste up to 200 gene symbols, then click Annotate Genes. Results appear as a sortable table with GO terms per gene. Download as CSV.

Data Sources

Accepted File Formats

Venn / Overlap — Input Format

Each list is a plain set of identifiers — gene symbols, Ensembl IDs, protein accessions, or any consistent identifier. Paste directly (one per line, or comma/semicolon/tab-separated), or upload a CSV/TSV/Excel file and select which column to use. Up to 10 lists, each up to any size. Lists are saved automatically between sessions.

DESeq2 / edgeR Output

For Volcano: upload the results table directly. Select log2FoldChange as the FC column and padj as the p-value column. Gene symbols are usually in the row index — if exported with row names, the first column contains them. For DESeq2 with Ensembl IDs, use the Gene Annotate tool to resolve to symbols first.

Volcano — Log Base Options

P-value vs Adjusted P-value

P-value — raw statistical probability. False positives accumulate when testing thousands of genes simultaneously.
Adjusted P-value (padj / FDR) — Benjamini-Hochberg correction for multiple testing. Most journals expect this for genome-wide analyses.

Heatmap — Clustering Methods

GO Plot — Dot Size Column

Single number — plain integer in the count column (DAVID, g:Profiler, Enrichr).
Gene list (semicolons) — symbols separated by / or ; (clusterProfiler). OmicsPlot counts symbols automatically.

Venn / Overlap — Guide

The Venn / Overlap tool compares gene lists across multiple experiments or comparisons. Enter up to 10 lists by pasting identifiers or uploading files. Lists are saved automatically — navigating away and returning restores your last session.

2–3 Lists: Venn Diagram

4–10 Lists: Similarity Network + Heatmap

Sending Lists to Venn from Other Tools

In Volcano, click → Venn to send Up or Down gene lists with context labels (e.g. "Up — MI vs Control"). In the Heatmap, send individual clusters. When you open the popup a second time with lists already loaded in Venn, you can choose Add to existing (append) or Start fresh (replace) — this is the key workflow for comparing DEGs across multiple comparisons.

Export Format Guide

Venn / Network Exports

Venn (2–3 lists): PNG at 3× resolution (≈288 DPI) or SVG via the Venn tab in the left panel.
Network (4+ lists): PNG or SVG via the Graph tab. Uses outerHTML capture — all colors are baked in, no CSS variable dependencies.
Gene Presence Table: Download CSV button inside the heatmap panel. Apply the "≥ N lists" filter first to limit to well-shared genes.
Pairwise Jaccard Matrix: Download Excel button in the matrix header — two sheets: human-readable (J + OC) and numeric-only for R/Python.

When should I use the Venn diagram vs the Network + Heatmap view?

Use the Venn diagram (2–3 lists) when you want a classic visual that reviewers immediately recognize and that clearly labels each intersection zone with a gene count. Use the Network + Heatmap (4+ lists) when you have more comparisons — Venn diagrams become geometrically impossible to read with 4+ sets. The network shows which comparisons are most similar to each other; the heatmap shows exactly which genes drive that similarity.

What is Jaccard index and why does it matter for comparing gene lists?

Jaccard index = shared genes / all unique genes across both lists. A value of 0 means no overlap; 1.0 means identical lists. It accounts for list size — two lists that share 50 genes but have 500 total unique genes score lower (0.10) than two lists sharing 50 genes out of 60 total (0.83). Use the Overlap Coefficient when one list is much smaller than the other, as Jaccard penalizes size differences.

Why is my heatmap blank or showing no data?

Most commonly because sample columns were left unassigned in the Groups step. Every column you want to display must be dragged into a group. Also check that columns contain numeric values — text, dashes, or empty cells may be excluded.

Why are genes missing from my volcano plot?

OmicsPlot removes rows with NaN in the FC or p-value column, and removes genes where Log₂FC = 0. You can uncheck these options on the column assignment page. If your FC column has very large values (e.g. 1000+), it is likely unlogged Raw FC rather than Log2.

My GO dots are all the same size. What's wrong?

Check the Dot size column mapping. If the column contains gene symbols separated by slashes or semicolons (clusterProfiler output), toggle to "Gene list (semicolons)" — otherwise it reads as text and sizing falls back to a uniform default.

What is k-cluster analysis and when should I use it?

Setting k ≥ 2 cuts the hierarchical dendrogram into k groups. Each group is color-coded. The main reason to use this is GO enrichment — clicking the GO → button on any cluster runs enrichment analysis on that cluster's genes. Start with k=3 and adjust based on biological sense.

Why does my export look different from the screen?

On-screen plots render at 96 DPI using your browser window size. Exports use your chosen physical dimensions and DPI (typically 300–600 DPI). Font sizes and margins are recalculated for export dimensions, so proportions may differ. Adjust font size in Typography settings if labels look too large or small in exports.

Does OmicsPlot store my data?

Uploaded files are stored temporarily in a session folder on the server for rendering only. Sessions expire and files are deleted automatically. Figure panels saved to Figure Builder are stored in your browser's local IndexedDB — they never leave your device. Venn lists are saved in your browser's localStorage. No account or login is required.