eMB: Mathematical Analysis of Cancer Evolution with New Sequencing Technology

University Of Minnesota-Twin Cities, Minneapolis MN

Investigators

Abstract

By sequencing the genome or transcriptome of a tumor, scientists and doctors can gain detailed insights into the tumor's genetic history. For example, how many mutations drove the tumor's progression, and what specific mutations were most important? Doctors can use this information to determine the best courses of treatment and the frequency of surveillance tests. In recent years, scientists have developed novel sequencing technologies -- including single-cell sequencing, multi-region sequencing, and sequencing of circulating tumor DNA (ctDNA). In this project, the investigators will develop new mathematical and statistical tools to analyze the large amounts of data generated by these new sequencing technologies. Their methods will leverage the latest knowledge of biological mechanisms driving cancer progression to develop constrained statistical models that maximize insights from cancer sequencing data. The investigators will incorporate real-world limitations of cancer sequencing data availability, e.g. limited time points and samples. This project will support the training of graduate students in inter-disciplinary science that combines mathematics, statistics, and bioinformatics. The mathematical tools will be built upon the underlying principles of population genetics which govern tumor progression. Due to the rapid expansion of tumor cell populations, work will primarily be done with branching process models; however, variants of the branching process model that incorporate finite carrying capacity will also be used, allowing for the accurate representation of real-world observations of tumor cell growth rates that decrease over time. Within the context of these stochastic models, the investigators will study several quantities related to new genomic sequencing technologies. For example, they will study the behavior of the site frequency spectrum in a branching process model with selection. This study will enable the development of tools to better detect the presence of so-called driver mutations in the genetic history of tumors. The investigators will also develop techniques based on multi-type branching processes for understanding data generated from multi-region sequencing. Additionally, techniques for utilizing single-cell sequencing to identify the mutations responsible for driving tumor progression will be developed. Finally, the investigators will study how ctDNA can be used to predict cancer recurrence and quantify the population dynamics of recurrent tumors. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.