Metabolic profiles to define the genome: can we hear the phenotypes?

Abstract

There is an increased reliance on genetically modified organisms as a functional genomic tool to elucidate the role of genes and their protein products. Despite this, many models do not express the expected phenotype thought to be associated with the gene or protein. There is thus an increased need to further define the phenotype resultant from a genetic modification to understand how the transcriptional or proteomic network may conspire to alter the expected phenotype. This is best typified by the description of the silent phenotype in genetic manipulations of yeast. High–resolution proton nuclear magnetic resonance ( 1 H NMR) spectroscopy provides an ideal mechanism for the profiling of metabolites within biofluids, tissue extracts or, with recent advances, intact tissues. These metabolic datasets can be readily mined using a range of pattern recognition techniques, including hierarchical cluster analysis, principal components analysis, partial least squares and neural networks, with the combined approach being termed metabolomics. This review describes the application of NMR–based metabolomics or metabonomics to genetic and chemical interventions in a number of different species, demonstrating the versatility of such an approach, as well as suggesting how it may be integrated with other ‘omic’ technologies.