Project 1 - Precision disparity modeling of cervical cancer survival using genomic and social determinants

点击翻译按钮获取中文摘要

ABSTRACT The National Research Council defines precision medicine as, “the ability to classify individuals into subpopulations that differ in their susceptibility to a particular disease, in the biology and/or prognosis of the diseases that they may develop, or in their specific treatment.” The underlying principle of this emerging discipline is to better capture and address key variability in health outcomes, giving emphasis to the role of genomic and epigenomic factors in disease etiology, onset, and progression. In the current proposal, we extend this principle to cancer health disparity. Our approach, termed “precision disparities” aims to better understand why certain population sub-groups persistently contribute to excess cancer risk. To accomplish this aim, we will examine multilevel risk factors, including genomic and epigenomic alterations, often not accounted for in most disparities-focused inquiry, as well as, explore the complex interaction between such alterations and socio-environmental risk conditions, associated with observed variability in cancer outcomes. For the current proposal, we will focus on cervical cancer survival. In the United States, this outcome is characterized by significant disparity. Blacks are more likely to die of this largely preventable disease, largely due to advanced stage of disease at diagnosis. However, there are some known biological differences in the cervical tumors of Black versus white women with disease that merit further inquiry, particularly in relation to modifiable behavioral and environmental determinants. Therefore, we propose: 1) to explore how DNA methylation and other multilevel risk factors moderate the influence of race on cervical cancer survival; 2) to identify patterns of disparity that are not race-based, but are driven by other underlying structure in the data; 3) to examine potential shared determinants, or drivers, across multiple disease phenotypes characterized by health disparity and, 4) to develop user-friendly software that will enable further disparities research. We will test such aims using the publically available Cancer Genome Atlas (TCGA), as well as BioVu and the Southern Community Cohort Study (SCCS) from Vanderbilt University. The latter two include patient socio-demographics ,clinical indicators, diet proxies, somatic mutation profiles, and community context, among other variables. The first three aims require the development of novel statistical methodologies, described in further detail in our grant text.