"The mutational landscape in cancers before and after treatment"Stein, AlexanderThe theoretical analysis of mutational patterns has a long-lasting history in population genetics and has been successfully applied to a variety of biological systems. In the past decade, a well-known quantity, the site frequency spectrum, has received considerable attention in the cancer evolution community, where it has been used to infer the evolutionary traits of cancers from sequencing data. Among other applications, the site frequency spectrum has been utilized to identify neutral evolution, quantify selection strength, and improve subclonal reconstruction. However, the potential of evolutionary insights within sequencing data remains largely unexplored. Here, we present the theoretical analysis of three quantities: (1) the site frequency spectrum, (2) the single-cell mutational burden distribution, and (3) the total mutational burden. We model the cell population using a birth-death process under the infinite-alleles assumption. Starting from a single progenitor cell, the population is growing to detection size, at which an abrupt change of growth leads to changes in the mutational landscape. We present analytic results for the above-mentioned quantities together with simulation results and discuss measurable signatures of treatment response in sequencing data. Interestingly, the total mutational burden and the single-cell mutational burden are scaling differently, which can be explained by the power law nature of the site frequency spectrum. In populations that remain approximately constant after an increasing phase, the total mutational burden is quickly dominated by newly arising mutations. Furthermore, constant populations lead to changes in the site frequency spectrum that may be interpreted as arising resistant subclone - if not considered carefully. In summary, we contribute to the theoretical development of modern population genetics and present its application to investigate treatment response in cancer. |
« back