I have broad interests in a number of cancer sites, but the underlying theme is identifying inter-individual differences in cancer susceptibility and using that to inform approaches to cancer prevention, early detection and precision medicine to enhance outcomes after diagnosis. Cancer reflects the interaction of genetic and environmental factors and observational research methods are needed to sort out this complex interplay. Importantly, genetic epidemiology is multidisciplinary, and I collaborate extensively with biostatisticians, bioinformaticists, molecular geneticists, biologists, pathologists, surgeons and radiologists.

For the past two decades my primary research focus has been on epithelial ovarian cancer, a rare disease but because it is diagnosed at late stage, the mortality rates are high. Numerous epidemiologic studies have identified non-genetic factors that influence risk, some that increase risk, and others that decrease risk. The disease also clusters in families, signaling the role of genetic predisposition. Early work focused on single nucleotide variants in genes involved in hormone metabolism (R01 CA86888), but with advances in technology that significantly decreased the cost of genotyping, efforts evolved to 'genome-wide association studies' (R01 CA114343; U19 CA148112) as part of an international collaboration: the Ovarian Cancer Association Consortium (OCAC). The U19 effort, in particular, represented efforts to link genetic variation identified through GWAS with functional changes. This grant was similar in spirit to a Program Project Grant, and brought together consortium of investigators who focused on cancers of the lung, breast, colon and prostate.

One of the realizations from the GWAS was that the vast majority of risk loci were not within coding regions of the genome. This required an entirely new strategy to evaluate functional effects, leveraging large extant databases curated by the NIH. It also led to the current focus on long non-coding RNA species (R01 CA207456). We have noted that there is a strong overlap in risk regions identified through GWAS with the physical location of lncRNA's in the genome, especially those that are active in ovarian tissue. 

As Cancer Center Director, 50% of my effort is dedicated to serving as PI on the Cancer Center Support Grant (CCSG)(P30-CA76292).

The balance of my research efforts are in collaboration with other faculty at Moffitt. These are described briefly under the collaboration tab.