It was nine a.m. on Wednesday, May 17^th and already the Ohio Union’s Interfaith room was nearly filled to capacity.  Perhaps drifting away from the room’s original purpose, the occupants had gathered for a tutorial on the field of metabolomics, or the study of the chemical components involved in life processes.  Expertise, or rather the lack thereof, was not a requirement for attendance, and thus the room was filled with a diverse array of folks from graduate students like myself to professors already specializing in metabolomics. 

As this session was the first event of the Joint Meeting of The Inaugural Conference on Food and Nutritional Metabolomics for Health and the 14th Annual Ohio Mass Spectrometry Symposium, the organizers were surprised at the high turnout to the tutorial.   While the training course focused on food metabolites, a lot of information given was relevant to many fields beyond food science. 

The three main topics presented at the tutorial were 1.) Study Design and Analytical Approaches, 2.) Pre-processing and Statistical Analysis, and 3.) Strategies for Metabolite Identification and Confirmation.  The session was concluded with a brief discussion on the future of metabolomics.  The methods discussed were largely for GC-MS or LC-MS.  Below are some of the take-away points:

Study Design and Analytical Approaches:

• To help reduce variability between replicates:
• The metabolite extraction process is the largest source of variability.  Try to use the same methods for all samples, even the same Lot # for solvents!
• Prepare samples as close as possible to when you run them (i.e. same/next day)
• Run all samples in a single batch if possible
• Unlike with targeted metabolite analysis, internal standards are not as helpful for untargeted analysis due to the diverse array of metabolites under consideration.  Usually, the internal standard represents a particular family or type of metabolite, such as amino acids or sugars, so that quantitation efforts are more accurate.  Therefore, it is not possible to reflect thousands of metabolites within a sample by one or even a few internal standards for untargeted analysis.

Pre-processing and Statistical Analysis:

• Many “features,” or peaks on the chromatogram, may actually be derivatives of a single compound.
• It’s important to look at your data, and not just depend on the computer software to identify peaks.

Strategies for Metabolite Identification and Confirmation

• Monoisotopic mass vs. nominal mass: those decimal points can be a big help!  The example given was C3H8 = 44.0624 and CH3CHO = 44.0261
• Once you have an accurate mass, you can search a plethora of databases.  This, however, can be time-consuming and overwhelming!  Try to narrow your list down by considering where the sample originated; some predictions just won’t make sense. 
• If you have MS/MS spectra, you can compare this to reference spectra.  Once you have a candidate metabolite, run a genuine standard (if available) on the same equipment and compare to your unknown.

The Future of Metabolomics:

• As more metabolomics studies are completed, there are a number of areas that need consideration:
• More standardization is needed across instruments, lab groups, databases, reporting/verbiage, data archival, etc.
• Right now it’s difficult to integrate data across the –omics.  In the future, it should be easier to combine datasets from a genomics experiment with metabolomics results, for example.

Online Resources:

Compound Identification Databases:

• metlin.scripps.edu https://metlin.scripps.edu/landing_page.php?pgcontent=mainPage
• The Human Metabolome Database: this one is interesting as you can filter the database to look for metabolites of plant origin.  A list of compounds is returned and shows where each compound was isolated (blood, saliva, urine, etc).  http://www.hmdb.ca/
• www.foodb.ca  : this one is a little more plant-centered as it lists compounds found in food crops.

NIH Synthesis Lab:

The NIH has allocated funding to create the Metabolite Standards Synthesis Center (MSSC), which is comprised of a number of university-institute partners that work together to provide researchers with genuine standards that are not commercially available.  More info can be found here at the Research Triangle Institute:  https://www.rti.org/impact/metabolite-standards-synthesis-center

Guidelines for reporting/describing metabolomics experiments:

http://metabolomicssociety.org/

According to their website, the Metabolomics Society was started in 2004 and now has over 1000 members from around the globe.  Their website has a long list of potentially useful databases for metabolomics studies, as well as software.  Additionally, there is information on how to best describe and standardize metabolomics studies for publication.  

Written by TPS Fellow, Irene Gentzel