Fig. 2

Analysis of 128 tumours reveal temporal dynamics and an effect of tumour type on population size. a-c Percentage of cells in healthy mammary fat pad (D0) and 4T1 and 4T07 tumours across the different time points. d-i Percentage of cells in healthy mammary fat pad (D0) and 4T1 and 4T07 tumours across the different time points positive for the respective CAF marker within the CAF+ population. Tukey box plots of two (healthy samples) or three independent repeats (tumour samples) combined. The boundaries of the Tukey style box go from the 25th to the 75th percentile, and the median is depicted by a line. The healthy samples provide a D0 reference point, but all the statistics are only concerning the tumour samples. Fitting a straight line to the tumour data, testing if one line fits both 4T1 and 4T07 tumours to indicate no difference, or if slope and/or y-intercept differ if two lines better capture the dataset. Fitted lines are shown with hashed out 95% confidence bands (CB95), with dots indicating the observed mean at the respective time points. a-c To determine statistical significance 2-way ANOVA with Tukey’s multiple comparisons post-test was used for intra-tumour comparisons, and multiple, two-tailed, t-tests with FDR correction (Q = 1%) were used for inter-tumour comparisons. d-i To determine statistical significance between time points within each tumour type (intra-tumour time comparisons) a 2-way ANOVA with Tukey’s multiple comparisons post-test was run on all the CAF markers combined, once for 4T1 tumours and once for 4T07 tumours. To determine if markers differed between tumour types, multiple unpaired, two-tailed t-tests without assuming equal variance and with FDR correction (Q = 1%) were conducted for each of the three time points (inter-tumour comparison).* = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** p < 0.0001. Healthy n = 12, 4 T1 D7 n = 21, 4 T1 D14 n = 24, 4 T1 D21 n = 24, 4 T07 D7 n = 24, 4 T07 n = 21, 4 T07 D21 n = 14