Labelling in metabolomics

[deeptijk]

We will discuss about labelling in Metabolomics here -

1. Should we make it mandatory/optional
2. Do we have CV terms to define the different types of labelling technique?

[deeptijk]

#703 (Metabolomics specification DRAFT [PLEASE DO NO MERGE])

[mmattano]

There is the use of labeled compounds as internal standards and the use of stable isotope tracers for isotopic labeling studies and, very rarely, the use of labeled isobaric tags for multiplexing.
Labeled compounds are frequently used as internal standards to enhance the accuracy and precision of quantitative analyses. These labeled standards, typically isotopically labeled (e.g., with stable isotopes like ^13C or ^15N), are chemically identical to the target analytes but can be differentiated based on their mass. By spiking a known quantity of these standards into a sample prior to analysis, researchers can correct for variability in sample preparation, extraction efficiency, and instrument response. This approach allows for more reliable quantitation of metabolites within complex biological samples, facilitating comparisons across different experimental conditions and contributing to more robust biomarker discovery and metabolic profiling.
For labeling studies it’s a bit more complex. In brief, we introduce stable isotopes by feeding it to a biological system as a substrate to introduce an observable variable into the system’s metabolism. There are three major use cases, either for tracing, labeling or to get input for flux calculations, or as I like to call them, where atoms come from, how they get there and how quickly they move around. Tracing tells us if the atoms of one metabolite are derived from the labeled substrate. For example, if you want to study if a specific metabolite is derived from a specific substrate, let’s say if TCA intermediates include atoms from circulating lactate, then you can introduce labeled lactate into your system, let some time pass, then take samples and measure the isotopologue distributions in TCA intermediates. If you detect a non-natural isotopologue distribution, then you can infer that your system must have incorporated labeled atoms from the labeled lactate. This is “tracing”, “labeling” would be that you look at the isotopologue distributions and try to learn more about how the labeled atoms must have moved through your system to produce the specific patterns you see. For example, if you feed system with fully labeled glucose, of which you should only measure the fully labeled isotopologue, but you measure half labeled isotopologues in glycolytic hexose intermediates, then you know that your system has gluconeogenetic activity, because stochastically you will have a recombination of labeled and unlabeled trioses during gluconeogenesis. And flux means that you use the measured labeling data to calculate how quickly reactions in your system must have happened to produce the observed patterns.
For multiplexing the use is similar as in proteomics; isobaric tags are used to derivatize samples and very few groups use differently labeled tags to multiplex samples. This is very rare though; I only know about Daniel Globisch and his group doing that.

[deeptijk]

@mmattano can it be mandatory field? In any case we must have a defined CV terms. If not then we have to add it. And do you have any example dataset?