PT - JOURNAL ARTICLE AU - Atsushi Niida AU - Satoshi Ito AU - Georg Tremmel AU - Seiya Imoto AU - Ryutaro Uchi AU - Yusuke Takahashi AU - Koshi Mimori AU - Satoru Miyano TI - Cancer evolution simulation identifies possible principles underlying intratumor heterogeneity AID - 10.1101/022806 DP - 2015 Jan 01 TA - bioRxiv PG - 022806 4099 - http://biorxiv.org/content/early/2015/07/20/022806.short 4100 - http://biorxiv.org/content/early/2015/07/20/022806.full AB - Cancer arises from accumulation of somatic mutations and accompanying evolutionary selection for growth advantage. During the evolutionary process, an ancestor clone branches into multiple clones, yielding intratumor heterogeneity. However, principles underlying intratumor heterogeneity have been poorly un-derstood. Here, to explore the principles, we built a cellular automaton model, termed the BEP model, which can reproduce the branching cancer evolution in silico. We then extensively searched for conditions leading to high intratu-mor heterogeneity by performing simulations with various parameter settings on a supercomputer. Our result suggests that multiple driver genes of moderate strength can shape subclonal structures by positive natural selection. Moreover, we found that high mutation rate and a stem cell hierarchy can contribute to extremely high intratumor heterogeneity, which is characterized by fractal patterns, through neutral evolution. Collectively, This study identified the possible principles underlying intratumor heterogeneity, which provide novel insights into the origin of cancer robustness and evolvability.