Figure 3—figure supplement 1.

(A) Flowchart of ISLA. To sample from intensity distributions, images are masked by setting voxels outside of the segmentation to zero and a simple background subtraction is performed. To sample from cell shapes, the 1vxl-wide outer shell of the segmentation is set to 1, all other voxels to zero. The resulting image is normalized and used to stochastically sample points for the point cloud. (B) Flowchart of CBE. Input point clouds of cells are either rotated according to a registration across tissues (Tissue Frame Of Reference, TFOR) or are volume-normalized and re-represented as a subset of the pairwise distances between points, removing size and rotational information (Cell Frame Of Reference, CFOR). A representative subset of the resulting clouds is overlaid and k-means clustering is performed on the overlay, yielding a set of common reference points. Finally, features are computed to describe each cell's point cloud relative to these common reference points, resulting in an embedded feature space. This feature space can be transformed with PCA to emphasize relevant variation across the sample population.