Time-resolved, integrated analysis of clonally evolving genomes

Clonal genome evolution is a key feature of asexually reproducing species and human cancer development. While many studies have described the landscapes of clonal genome evolution in cancer, few determine the underlying evolutionary parameters from molecular data, and even fewer integrate theory with data. We derived theoretical results linking mutation rate, time, expansion dynamics, and biological/clinical parameters. Subsequently, we inferred time-resolved estimates of evolutionary parameters from mutation accumulation, mutational signatures and selection. We then applied this framework to predict the time of speciation of the marbled crayfish, an enigmatic, globally invasive parthenogenetic freshwater crayfish. The results predict that speciation occurred between 1986 and 1990, which is consistent with biological records. We also used our framework to analyze whole-genome sequencing datasets from primary and relapsed glioblastoma, an aggressive brain tumor. The results identified evolutionary subgroups and showed that tumor cell survival could be inferred from genomic data that was generated during the resection of the primary tumor. In conclusion, our framework allowed a time-resolved, integrated analysis of key parameters in clonally evolving genomes, and provided novel insights into the evolutionary age of marbled crayfish and the progression of glioblastoma. Genomes evolve under the accumulation of mutations, and under the pressure of selective forces. While additional mechanisms are at play in sexually reproducing species, this is not the case in clonal genomes. Our study focuses on a parthogenetic animal and on cancer, since both possess a clonal genome, and in both cases evolutionary forces are key to understand expansion. We used modelling of mutation accumulation, in combination with Darwinian selection and with clock-like mutagenic processes. Using this framework, we showed a remarkably recent emergence date for P. virginalis and established its potential as a model system for clonal genome evolution. We highlighted subtle temporal dynamics of selection in tumor samples, and showed that tumor cell survival was correlated with the time to recurrence. Our findings illustrate the potential of this framework for modelling of clonal evolution and for the use of evolutionary parameters in a clinical context.