Fig. 2: Spatiotemporal dynamics of microglial polyclonal proliferation after ischemic stroke revealed by Monte Carlo simulation and machine learning.

a Representative 3D renderings of Confetti⁺ microglia in matched ischemic and contralateral striatum. Image dimensions are presented in µm. b Illustration of the applied Monte Carlo simulation method. Densities of same-colored cells were calculated within concentric rings with radii ranging from 20 to 300 µm and a fixed width of 20 µm. The enlarged image shows an example ring with a radius r = 50 µm. Hard lines represent recorded data, while colored regions represent randomized simulated data. The upper and lower bounds of simulated data correspond to the 98th and 2nd percentiles, respectively. c Time course analysis with Monte Carlo simulations for microglial post-stroke polyclonal proliferation. The polyclonal proliferation of microglia is present 2 days after stroke. The recorded cell densities reach their peak after 2 weeks and shift back toward random distribution after 12 weeks. The unaffected, corresponding region in the contralateral hemisphere shows no signs of clonal dynamics. d Cluster analysis with DBSCAN for an exemplary image from the GFP channel by employing a neighborhood radius (ε) of 50 µm. Cells that are further than 50 µm from their closest neighbor are considered singlets and were excluded from the clonal analysis in (e, f). Different colors stand for different clones. Scale bar: 100 µm. e The dynamics of the polyclonal proliferation of microglia represented by the number of clones of Confetti⁺ microglia in ischemic and contralateral striatum for each investigated time point, calculated by DBSCAN. f The dynamics of changes in the clone size over time. In (e, f), each dot represents one brain hemisphere from one mouse. Means ± s.e.m are shown. Statistical analysis with two-way ANOVA, Šidák’s multiple comparisons test for contralateral vs. stroke comparisons, and Dunnett’s multiple comparisons test for time point comparisons. Number of animals (the same animals used in Fig. 1) N = 8 (2 d), 8 (1w), 7 (2w), 7 (4w), 7 (8w), and 6 (12w).