Simulation Framework for Generating Intratumor Heterogeneity Patterns in a Cancer Cell Population

Abstract

As cancer cell populations evolve, they accumulate a number of somatic mutations, resulting in heterogeneous subclones in the final tumor. Understanding the mechanisms that produce intratumor heterogeneity (ITH) is important for selecting the best treatment. Although some studies have involved ITH simulations, their model settings differed substantially. Thus, only limited conditions were explored in each. Herein, we developed a general framework for simulating ITH patterns and a simulator (tumopp). Tumopp offers many setting options so that simulations can be carried out under various settings. Setting options include how the cell division rate is determined, how daughter cells are placed, and how driver mutations are treated. Furthermore, to account for the cell cycle, we introduced a gamma function for the waiting time involved in cell division. Tumopp also allows simulations in a hexagonal lattice, in addition to a regular lattice that has been used in previous simulation studies. A hexagonal lattice produces a more biologically reasonable space than a regular lattice. Using tumopp, we investigated how model settings affect the growth curve and ITH pattern. It was found that, even under neutrality (with no driver mutations), tumopp produced dramatically variable patterns of ITH and tumor morphology, from tumors in which cells with different genetic background are well intermixed to irregular shapes of tumors with a cluster of closely related cells. This result suggests a caveat in analyzing ITH data with simulations with limited settings, and tumopp will be useful to explore ITH patterns in various conditions.\n\nAuthor SummaryUnderstanding the mechanisms that produce intratumor heterogeneity (ITH) is important for selecting the best treatment. Despite a growing body of data and tools for analyzing ITH, the spatial structure and its evolution are poorly understood because of the lack of well established theoretical framework. Herein, we provide a general framework for simulating ITH patterns, under which a simulator (tumopp) is developed. Tumopp offers many setting options so that simulations can be carried out under various settings. Simulations using tumopp demonstrate that dramatically variable patterns of ITH and tumor morphology can be produced depending on the model setting. The present work provides a guideline for future simulation studies of cancer cell populations.