Fig. 4.

Stem cell dynamics are accelerated within Ctnnb1-driven tumours but inhibited adjacently to them. (A) Schematic diagram and real example of measurements used to characterise clones inside and outside of tumours. (B) Graph displaying average clone size across binned median of tumour sizes. Each bin contains 50 clones. Points and bars show median and interquartile range. (C) Graph comparing the proportion of surviving clones that have monoclonally converged within tumours and in surrounding tissue. (D) As in C but with the intra-tumour data split into those clones that are further from (tumour edge) or closer to the Tc (falling without or within 75% of the average tumour radius, respectively). (E) The proportion of surviving non-fixed clones (in crypt quartiles) inside and outside tumours evolve differently in time when the monoclonal proportion inside tumours has been rescaled to be identical to the monoclonal proportion outside tumours. (F) Divergence from theoretical wild type of the clonal dynamics in crypts occupying adjacent intestinal epithelia surrounding β-catenin tumours. Clones are grouped into overlapping bins containing equal numbers of crypts (100 clones per bin) increasing in distance from the edge of the tumour. Line and ribbon shading show the 80% and 95% credible intervals around the median of 1000 simulated datasets of 100 wild-type crypts. Point sizes show the average clone size in each bin. Distances are measured in crypt diameters. (G) The daily replacement rate λ of stem cells undergoing neutral drift is inferred for clones in tumour-adjacent epithelia. Close to tumour (Re<7 crypt diameters), λ=0.07 [80% confidence interval (c.i.): 0.061-0.083]; far from tumour (Re≥7 crypt diameters), λ=0.09 (80% c.i.: 0.079-0.103).