Fig. 4

Effect of sample size on clonality inference in spatial tumour. To test the effect of biopsy size, we tested the accuracy of clonality estimations by sampling groups of cells (one sample = 200 cells, corresponding to 1% of tumour size). Batches of tumour samples (two for a), five for b) and ten for c)) were taken at random locations and used to estimate the clonality of the present mutations. We repeated sampling process 10 000 times for on each tumour and calculated proportion of correct estimations – single point on figure. We exclude mutations below a certain frequency ε from the analysis (ε=0.0 or no exclusion for red, ε=0.3 for blue, ε=0.8 for green), which increases the accuracy of clonality estimations