header photo

Marissa Kosnik successfully defends her PhD!

Title - Data integration elucidates connections between environmental factors and health outcomes
Abstract : It is well understood that environmental factors contribute to the development of human diseases. Multiple types and sources of data are necessary to address the complex relationship between public health and environmental exposures. Identification and characterization of these relationships requires analyzing data from multiple biological levels, from molecular studies of chemical-gene perturbations through epidemiological studies of population responses, to align results in a human health framework. Numerous databases have been developed to elucidate the role of these factors and diseases, as new data is continually generated. These databases cover a variety of different data levels and types, from high-throughput screening (HTS) efforts like the Tox21/ToxCast initiative through human health data from sources like the U.S. Renal Data System (USRDS). However, forming connections between these different data poses challenges in both implementation and interpretation. Challenges in data interpretation include alignment of in vitro toxicity results in a biological context, interpretation of chemical-endpoint perturbations, and alignment of human health data in an exposure framework. Finding ways to connect these data through common data elements is crucial in building connections between environmental factors and human health outcomes. Chapter 1 is an overview of different publicly available data sources of environmental factors and human health outcomes and describes these data in the context of Chapters 2-5. Chapter 2 describes an analysis of HTS data from micro-electrode arrays to develop an assay for chemical neuroactivity and subsequent characterization of the chemical-endpoint results. Chapter 3 demonstrates the utility of data integration to provide biological context to HTS results through an analysis of Tox21/ToxCast data integration with the Comparative Toxicogenomics Database. Chapter 4 characterizes chemical-disease associations in a risk context by integrating chemical-gene, gene-disease, and variant-disease data to develop new risk values to prioritize associations for improved human health risk assessment. Chapter 5 applies data integration techniques in a public health context through survival analysis for patients with end-stage renal disease utilizing patient data from the USRDS, environmental quality data, and geospatial information on hospital locations. Chapter 6 summarizes data integration techniques and describes future directions.

Go Back