here.">

Created by: Nathan Lee

Issue 236: This is an abstract visual representation of a Monte Carlo simulation of carcinogenesis, similar to simulations we used in our recent paper where we reconstruct the evolutionary history of liquid cancers, including when cancer was initiated and when subsequent driver mutations occurred. Before applying our methods to clinical data, we evaluated them on simulated cancers. To generate this image, I perform a Monte Carlo simulation with birth, death, and mutation to generate a population of 600 cells. The first cell starts with no mutations, and its wild-type state is represented by a square. Upon each mutation, I sample from a distribution to determine that square’s "genotype," or the noise added to each corner. The effect of new mutations is added to the existing genotype of each square. The opacity of each polygon is determined by the total number of mutations accumulated by the cell. To learn more, check out the paper here.