): The cell lineages in neoplastic tissues are evolving under the twin dynamics of mutation and clonal expansion. This is the basis for both its virulence and our difficulties in treating it. We hypothesize that a subset of the mutations observed in the progression to cancer confer beneficial selective effects on the cell. Furthermore, we hypothesize that the interactions between these clonal populations in and around the neoplastic tissue determine the progression to cancer.
The aim of this project is to identify these selective mutations and to infer the interactions between the mutant clones in Barrett's Esophagus that eventually lead to the development of esophageal adenocarcinoma. This analysis will be based on loss of heterozygosity data, promoter methylation, single nucleotide polymorphisms, and gene expression data for homogeneous subpopulations of cells sampled from neoplastic tissue in patients with Barrett's Esophagus. The tissue samples come from biopsies of selected patients in the Seattle Barrett's Esophagus Project (N=285). Data mining will be used to identify mutations that are associated with clonal expansion as well as inhibition of neighboring clones. The order of genetic events in the progression to esophageal adenocarcinoma will be determined by phylogenetic reconstruction of the cell lineages in the neoplastic tissue of each patient. Machine learning techniques, such as the EM algorithm, will be employed to infer missing data and the effects of unsampled mutations. Computational modeling will be used to generate comparison data for null hypotheses as well as to generate experimental predictions from our understanding of the progression to cancer. This is the first step in my long-term career goal to contribute to medicine through the use of computational and theoretical methods. Working at the Fred Hutchinson Cancer Research Center will facilitate the transition from my background in computer science and evolutionary theory, to an independent research program based on the analysis of cellular and molecular dynamics of cancer. The challenge of this project is the integration of diverse molecular and epidemiological data into a coherent and detailed understanding of the progression to cancer in the neoplasm of a model system.
Showing the most recent 10 out of 11 publications
