Center for Metabolomic Studies Seminar Series: “Standardization of Nontargeted Metabolomics Improves Interpretation of Biomedical Studies”
Oliver Fiehn, PhD
University of California, Davis Genome Center – Metabolomics
Presentation: “Standardization of Nontargeted Metabolomics Improves Interpretation of Biomedical Studies”
Nontargeted metabolomics combines quantifying many classic metabolites with detecting and identifying novel compounds. Nontargeted metabolomics aims to generate new hypotheses. A range of different techniques and metabolite names are used between academic and commercial providers that make it difficult for biomedical scientists to compare results between studies, or even to interpret findings. Conversely, targeted metabolomics focuses on known metabolites. Targeted metabolomics is ideal for quantitative comparisons between studies and between laboratories, but focuses on validating specific hypotheses.
How can these worlds be unified? We here present a range of advances. First, kits of internal standards enable absolute (molar) quantifications of metabolites in nontargeted screens by combining adducts across organs. Second, novel instrumentation, from the IQX- and Ascend-type Orbitraps to electron activated dissociation, gives us new capabilities to detect low abundant target metabolites during non-targeted analyses, as well as better defining positional isomers in natural products and lipids. Third, new cheminformatics approaches improve confidence in compound annotations, from spectral entropy similarity to LibGen spectral denoising for low abundant compounds. Most importantly, kits of internal standards can standardize non-targeted metabolomics into (international) databases, as we show by examples from the ‘unknown lipids’ metabolomics working group, the PTFI global food initiative and the new UC Davis LC-BinBase database.
Overall, application of these advances will be exemplified by studies on microbiome research in the human small intestine, exposome identifications across human plasma samples, and mouse phenotyping.