Lipid Analysis for Complex Communities
Non-linear transformation of phospholipid fatty acids (PLFA) profiles and feature extraction find patterns of exposure response in a freshwater periphyton community.
Phospholipid fatty acids (PLFA) have been widely used to characterize environmental microbial communities, generating community profiles that can distinguish phylogenetic or functional microbial groups within the community at a very high-level.
A periphyton community was exposed to uranium at four doses, (1) un-amended, (2) 10 ug/L, (3) 100 ug/L, and (4) 500 ug/L over a course of 5 days. Following collection of the periphyton the biomass was extracted and assayed for PLFA. A non-linear statistical learning methodology permitted a more robust biological interpretation of the PLFA profiles in terms of the types of organisms that were affected by uranium exposure and those that were not affected. In addition, the in-depth analyses into the data structure via kernel mappings formed the foundation for a probability model that classifies a sample into a uranium exposure category based on the PLFA profile alone. The results of these statistical analyses identified changes in the periphyton community in response to the uranium that other, more commonly-used statistical methods were not able to identify.
Contact: Amoret Bunn