How to create a multi-label characteristic group from exclusive classifications

This guide shows you how to assemble multiple one-hot-encoded classifications into a multi-label characteristic group where each state combination is a phenotype.

Prerequisite: the segmentation and measurements for at least one population.

Reference keys: single-cell measurement, characteristic group, phenotype

Step-by-step:

1. Perform successive conditional binary classifications to isolate your phenotypes of interest from the rest. For each phenotype, classify the phenotype vs all other cells.

2. Click on the Explore table button to open the table view.

3. Select your groups or status classification-like features to merge them into a single characteristic group by Ctrl+Left-Clicking on the column to combine.

4. Go to Math > Merge states…. Set a name for the new group. Check the feature names to merge.

5. Compute. The new column is added add the end of the table view.

6. Go to File > Save inplace… to write this new feature in all of the position tables.

7. Enter the viewer in the MEASURE section to explore the new group.