Cluster Heatmap

This page is located at Visualize > Cluster Heatmap.

Annotation Cluster Heatmap Page

On-screen Items

a. – Summary Selection

The summary selection field is where you choose a summary CSV file generated by Palmreader via the Analyze > Analyze page with the “Summary CSV” option enabled. Pressing the icon will open the file picker, allowing you to select the CSV file.

b. – Select Rows Button

The row selection system is shared between several visualizations pages and is documented in the Shared Items page.

c. – Select Columns Button

The column selection system is shared between several visualizations pages and is documented in the Shared Items page.

d. – Grouping Variable Selection

The grouping variable selection field allows you to choose which variable to group your subjects with. The list of possible values depends on the variables used in the recordings summarized in the selected CSV file.

e. – Grouping Mode Selection

The grouping mode determines how the Z-scores will be computed and used to group the subjects.

“Group” – Group-level Z-scores

In “Group” mode, the Z-scores of individual animals within each group of the selected variable will be averaged. This mode will provide significance markers on those features which are most significantly different between the groups. The significance markers take the form of the text *, **, and *** overlayed on the rows representing the features with the significant difference.

This mode utilizes ANOVA-based statistical testing with Bonferroni-corrected p-values to highlight the significance.

“Individual” – Individual-level Z-scores

In “Individual” mode, the Z-scores of individuals animals are used directly. The subjects are clustered independent of groups, allowing you to uncover latent cluster structures within the dataset. Once the clustering is applied, the group identity is recovered and displayed, allowing you to asses whether these underlying structres align with predefined groupings. This mode is useful for identifying patterns and verifying if natural data clusters align with existing group labels.

f. – Generate Graph Button

The generate graph button is shared between several visualizations pages and is documented in the Shared Items page.

g. – Save Graph Button

The save graph button is shared between several visualizations pages and is documented in the Shared Items page.

h. – Graph Frame

The save graph button is shared between several visualizations pages and is documented in the Shared Items page.