Tab Gene Correlation
The Correlation tab computes Pearson and Spearman correlation between the expression of two selected genes across cells, with optional normalisation and cluster filtering, and displays a scatter plot alongside the computed statistics.
| Control | Description |
|---|---|
| Gene A | First gene to correlate. Defaults to the first gene in the panel. |
| Gene B | Second gene to correlate. Defaults to the second gene in the panel. |
| Normalisation | How to prepare expression values before computing correlation: Raw counts, Fraction of total, or Log1p(CPM). Defaults to Log1p(CPM). |
| Filter by current cluster selection | When checked, restricts the analysis to cells currently visible in the Cell Coloring filter |
| Plot Correlation | Computes Pearson r and Spearman rho, generates a scatter plot with a statistics box, writes the figure to disk, and displays correlation values in the status bar |
- Select Gene A and Gene B from their respective dropdowns.
- Choose a Normalisation method appropriate for your comparison.
- Optionally enable Filter by current cluster selection to restrict the analysis to a cell subset defined in the Cell Coloring tab.
- Click Plot Correlation to generate the figure and read off the statistics.
- At least two genes must be present in the dataset for this tab to be functional.
- If cluster filtering is enabled but no filter is active in Cell Coloring, all cells are used.
- The saved figure includes both correlation coefficients and the number of cells used.
-
Plot Correlation does not open a save dialog. It writes to
<dataset>/plots/gene_correlation.<fmt>every time, where the format follows Preferences → Plot format (SVG by default, PNG at 300 dpi if selected). Each run overwrites the previous file, so rename anything you want to keep. -
Log1p(CPM)uses the shared normalised expression data, computing it once if no other tab has yet; the other two normalisations read the raw counts directly.