ERK activity in starved MCF10A WT, and KRAS G12V and PIK3CA H1047R mutant cell lines with and without the MMP inhibitor batimastat. (A) Population-averaged ERK activity at 10 h after the start of acquisition. Box plots show the median and 25th and 75th percentiles, whiskers correspond to minimum and maximum non-outlier values; N ≈ 2,000 cells pooled from three FOVs per cell line/treatment of a single experiment. Symbols indicate the statistical significance of a two-sample Wilcoxon rank sum test with Bonferroni correction: * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001. Numbers indicate Cohen’s effect size d (Cohen, 2013). (B) A random sample of single-cell ERKATS from panel A. (C) Frequency of ERK activity pulses per cell for ERKATS from panels A and B. A pulse is a continuous period of binarized high ERK activity. Symbols and numbers as in A. (D) X-axis projections of single cells during their participation in CEs. Events are colored with a finite discrete palette; thus, colors repeat over time. Data from single FOVs. (E) Size and duration of CEs. Hexagonal tiles are colored according to the number of CEs (logarithmic scale). Red dots indicate median statistics. Data pooled from three FOVs per cell line/treatment of a single experiment. (F) Results of statistical tests for the size and duration of CEs for the data in E. Symbols and numbers as in A. (G) Each dot corresponds to the mean number of CEs in a FOV normalized to the time-averaged number of cells in that FOV. The results of unpaired two-sample t tests with Bonferroni correction are shown above the plot. Symbols and numbers as in A. (H) Global Moran’s I calculated over time (287 time points) for each FOV. Each distribution contains values pooled from three FOVs and is based on N = 3 × 287 values. Vertical dashed lines represent the median. P values obtained from the Wilcoxon rank sum test; d, Cohen’s effect size.