Fig. 1

Spatial model of intratumour heterogeneity. a Schematic illustration of our spatial cancer model. Tumour cells are represented as nodes on a two dimensional lattice. Each cell has a mutational profile. With each cell division, a cell can mutate with probability μ=0.5 and thus intratumour heterogeneity is generated. While we typically think of each node as a cell, it could in principle also represent a small subpopulation of cells. b The mutational profiles of cells within a bulk biopsy are then combined to give the mutational profile of the biopsy sample. Red coloured subpopulations are cells that carry the first sub-clonal mutation (red in the mutational profile), while blue cells do not contain it. All cells carry one truncal mutation (blue in the mutational profile). Cells acquire new mutations and are presented with a different nuances of the original colour. In this example, because the red and blue subpopulations are approximately equal in size, we call the tumour well-balanced. c The different first-tier sub-clones are spatially represented within the tumour, placed around the centrally positioned ancestral population. d An ancestral population that contains a set of truncal mutations (blue square) branches multiple times into first-tier sub-clones with their own private sets of mutations. Having only samples from one of the major branches will result in a misclassification of the sub-clonal mutation that founded that branch