Technology
Metabolomic analysis is a process in four steps:
I) Extraction
II) Separation and detection
III) Bioinformatics
IV) Biomarker identification and data interpretation
The extraction is a fundamental step, which decides on the coverage and yield of metabolites. This step is also the most senstive and thus needs to be strictly controlled. Depending on the method of extraction different compound classes will be enriched or discriminated.
In most metabolite profiling methods the compounds will be separated due to their charge and size in most cases employing gas or liquid chromatography. Due to the large variation in physical properties of metabolites different separation technologies need to be employed.
When the metabolites are separated and leave the column they are injected in to the mass spectrometer, where the detection occurs. The compound will be detected as a whole molecule or fragemented using electrons. The fragmentation pattern is very specific for each compound and thus can be matched with a database of known compounds.
For a metabolic fingerprinting approach the extract can also be directly infused into the mass spectrometer. This allows the detection of differences in compounds and their concentration, but makes metabolite identification more difficult.
Unknown metabolites present in samples can be identified using ultra sensitive and high-resolution mass spectrometry such as FT-ICR-MS systems.
The complexity and wealth of information retrieved in a single experiment renders solid bioinformatic support necessary. Only with the help of proprietary software, specific algorithms and custom-build databases can such datasets be evaluated to the full and meaningful data generated.
IV) Biomarker and metabolite QTL Discovery
The final step is the evaluation of the dataset within the biological context and is performed by highly-skilled scientists. They will extract the information and determine whether the dataset can be categorized in different groups of biological samples. Find metabolites, so called biomarkers, which can predict certain phenotypes or traits or whether certain samples have significantly different metabolite concentrations and thus are delineated to be metabolite QTLs.