Fig. 8: Spatial interaction maps of integrated cellular structures.

A Each correlation heatmap shows the pairwise interaction between intracellular structures within each timepoint and treatment condition. Pearson correlation coefficient (PCC) was used to measure the spatial interaction of morphed signals of the intracellular structures after cell alignment with hexagon-based template matching. A clustering algorithm was applied to each triangle heatmap to generate dendrograms displaying the distance between clusters. A color code based on structure location was assigned to each intracellular structure. B Serial network evolution graph showing changes in organelle connectivity with time. Line color indicates PCC values, while line width corresponds to absolute values. C Network graph representing the difference between PCC values at week 4 and week 1 with PGE2 (left) and HPI4 (right). D Line graph displaying organelles weighted degree of centrality of PGE2-treated cells across weeks. Colored lines are the top three organelles with the highest average value. E Line graph showing the relative change of weighted degree to week 1 for PGE2 and F HPI4. G Dot plot shows the first week of significant change in interaction. The color indicates PCC difference to week 1, and dot size displays the absolute change. H Organelle-specific features were extracted from Random Forest predictions and ranked by importance. I Heatmap showing feature importance score weighted by model accuracy.