Environmental physiology, toxicology, and ecology and evolution stand to benefit substantially from the relatively recent surge of “omics” technologies into these fields. These approaches, and proteomics in particular, promise to elucidate novel and integrative functional responses of organisms to diverse environmental challenges, over a variety of time scales and at different levels of organization. However, application of proteomics to environmental questions suffers from several challenges—some unique to high-throughput technologies and some relevant to many related fields—that may confound downstream biological interpretation of the data. I explore three of these challenges in environmental proteomics, emphasizing the dependence of biological conclusions on (1) the specific experimental context, (2) the choice of statistical analytical methods, and (3) the degree of proteome coverage and protein identification rates, both of which tend to be much less than 100% (i.e., analytical incompleteness). I use both a review of recent publications and data generated from my previous and ongoing proteomics studies of coastal marine animals to examine the causes and consequences of these challenges, in one case analyzing the same multivariate proteomics data set using 29 different combinations of statistical techniques common in the literature. Although some of the identified issues await further critical assessment and debate, when possible I offer suggestions for meeting these three challenges.
© The Author 2012. Published by Oxford University Press on behalf of the Society for Integrative and Comparative Biology. All rights reserved.
W. Wesley Dowd; Challenges for Biological Interpretation of Environmental Proteomics Data in Non-model Organisms. Integr Comp Biol 2012; 52 (5): 705-720. doi: 10.1093/icb/ics093