Algorithms for Inferring Tumor Composition and Evolutionary History
October 30, 2017 - 11:15am to 12:15pm
Keller Hall 3-180
Abstract: Cancer is a disease resulting from somatic mutations - those that occur during the individual’s lifetime - and cause the uncontrolled growth of a collection of cells into a tumor. Recent progress in DNA sequencing technologies have enabled the development of personalized approaches to medicine where a patient’s treatment may be tailored to their specific genomic architecture. However, such approaches require accurate identification and interpretation of the set of mutations within each patient’s genome: still a difficult task, especially for cancer genomes. For example, tumors are the result of an evolutionary process where the temporal acquisition of mutations lead to multiple distinct tumor cell populations - each with their own complement of genomic mutations - within a single tumor. DNA sequencing data often represents a mixture of these populations, rather than any individual population. In this talk, I will describe several algorithmic methods that are able to deconvolve this type of mixed DNA sequence data to infer the original set of tumor populations and reconstruct the evolutionary history of the tumor.
Bio: Layla Oesper is an Assistant Professor of Computer Science at Carleton College. Her research interests lie in the development of algorithms and mathematical models to aid in answering important biological questions. Specifically, her recent work has focused on the design of algorithms for analysis and interpretation of high-throughput DNA sequencing data of cancer genomes. Dr. Oesper received her B.A. in mathematics from Pomona College and her Sc.M and Ph.D. in Computer Science from Brown University. She is a Google Anita Borg Scholar and is the recipient of an NSF Graduate Research Fellowship and an NSF CRII Award.