Self-Organizing Map Approach to Characterizing Complex Environmental Mixtures for Environmental Health Research
The NIEHS Exposure Science and the Exposome Webinar Series
April 27, 2017
Recordings of past events are available on YouTube - Exposome.
Join Webinar LISTSERV
Register for the Webinar LISTSERV to receive information on upcoming NIEHS Exposome webinars.
Literature on Exposome
The NIEHS Strategic Plan places a significant emphasis on transforming exposure science through the development of new approaches to exposure assessment, the definition and dissemination of the exposome concept, and the development and demonstration of the exposome as a tool for both epidemiological and mechanistic research. In order to achieve this goal, NIEHS launched the Exposure Science and the Exposome Webinar Series on April 4, 2014 to foster discussions on international efforts in advancing exposure science and the exposome concept as well as challenges and opportunities in incorporating this concept in environmental health research.
The complexity and dynamic nature of environmental pollution creates many challenges for health investigators seeking to illuminate health effects involving exposure to complex environmental mixtures. Identifying relevant mixtures, defining which are most important, and estimating health effects are some of the various challenges presented by this area of research. This talk will introduce the 'environmental mixtures problem' and then describe an approach that was found useful for research in this area -- the self-organizing map (SOM). This will entail a description of the SOM algorithm and its application to develop 'mixture' classification systems that can support research involving complex environmental mixtures. Application of the SOM will be illustrated using a variety of environmental data sets -- highlighting both benefits and limitations. Finally, this talk will demonstrate how to use SOM results epidemiologically in the context of an acute health effects study of air pollution. At the end of this talk viewers will have a better appreciation of the challenges presented by complex environmental data and become aware of an appealing tool to add into their 'mixtures' toolkit.
John L. Pearce, Ph.D., is an Assistant Professor of Environmental Health at the Medical University of South Carolina. He is the leader of the MUSC Air Quality Exposure Lab and serves as principal investigator for an R00 awarded by the National Institute of Environmental Health Sciences that focuses on developing approaches to better understand and assess complex environmental exposures. Dr. Pearce has published work on exposure monitoring, modeling, and characterization. His long-term research goal is to pioneer methodologies that will lead to breakthroughs in the understanding of links between complex environmental exposure and human health. Dr. Pearce holds a Ph.D. in Geography and Environmental Science from Monash University in Melbourne, Australia and completed his postdoctoral training in environmental biostatistics at Emory University in Atlanta, GA conducting research with the Southeastern Center for Air Pollution Epidemiology, a USEPA Clean Air Research Center.