Fig. 2: Growth modes impact the extent of clonal diversification and tumour fitness.
a, Schematic of the whole-tumour analysis of clonal diversity. b, Heatmap showing the average number of clones (that is, parental clone and subclones) with respect to driver acquisition probability (pdriver) and selective coefficient (s) in the volume growth (i) and surface growth (ii) models. The average is calculated from 50 in silico tumours per parameter condition. Clones with a whole-tumour CCF of at least 0.05 are counted. c, Whole-tumour CCF of parental and largest subclones in in silico tumours under volume growth (i,ii) and surface growth (iii), respectively. Average fitness in a tumour slice for each simulation is presented as a heatmap. Driver acquisition probabilities in these sets of simulations are pdriver = 2 × 10−4 (i), 1 × 10−3 (ii) and 2 × 10−4 (iii). ‘Parental (3p loss, VHL)’ clone is shown along with up to five subclones with a whole-tumour CCF of 0.01 or higher. All remaining subclones are represented in the ‘other’ group. d, Whole-tumour CCF of parental clone in in silico tumours under volume growth and surface growth with varying driver acquisition probabilities. n = 100 for each condition. e, Shannon diversity index in in silico tumours under volume growth and surface growth with varying driver acquisition probabilities. n = 100 for each condition. f,g, Mean fitness of randomly sampled (10% of all) tumour voxels against the mean fitness of the central-most (10%) tumour voxels, in models with saturated (f) and additive (g) driver advantages. Data points reflect sets of simulations with varying growth patterns (colour), driver acquisition rates (size) and implementation of necrosis (symbol). Heatmaps indicate the fitness in representative in silico tumours under surface growth without or with the implementation of necrosis. Statistical annotations in d and e reflect two-sided Wilcoxon tests: **** P ≤ 0.0001. In box plots in d and e, the ends of the box reflect the lower (Q1) and upper (Q3) quartiles, with the difference indicating the IQR; the horizontal line dividing the box reflects the median; the ends of the vertical line indicate the extreme values within the range from \({\mathrm{Q}}1 - 1.5 \times {\mathrm{IQR}}\) to \({\mathrm{Q}}3 + 1.5 \times {\mathrm{IQR}}\); dots beyond the vertical line indicate potential outliers.
