Researchers from the Scripps Research Institute and Washington University, St Louis, have published a paper in Nature Biotechnology that describes recent expansions to the Metabolite Link, or METLIN, database as well as potential benefits for the field of metabolomics.
METLIN is an online repository that provides tools for archiving, visualizing, and analyzing metabolite and tandem mass spectrometry data. The Nature Biotech paper highlights two developments in the current version of the resource: a new informatics workflow that automates the process of identifying metabolites, and an increase in the quantity of information contained in the database.
In the current version of the database, investigators can upload mass spec data directly into METLIN, and the system automatically searches for matches between the data from the research samples and information in the repository's library, the paper explains. METLIN measures the quality of the matches by using a scoring system based on the X-Rank algorithm that was developed by researchers at the Swiss Institute of Bioinformatics.
In previous incarnations of the database, this data-matching step was performed with a workflow that required seven manual steps. METLIN's automated procedure, however, could cut that number down to two steps, the paper explains.
Without this automated ability, "the value of metabolomic data would be very limited,"Gary Siuzdak, senior director of Scripps Center for Metabolomics and Mass Spectrometry and one of METLIN’s initial developers, said in a statement.
That's because "each study would require manual searches and ultimately fail to culminate in enough reference data to arrive at conclusive metabolite identifications," he explained.
METLIN's automated workflow should cut the time needed to "perform untargeted profiling and the subsequent metabolite identification" down to "minutes or hours as compared to the days or weeks needed with the traditional workflow," the authors wrote.
Furthermore, METLIN's developers have added high-resolution MS/MS spectra data on more than 10,000 metabolites contained in the database, the paper states.
METLIN also gives users access to more "physiologically relevant" metabolites — or metabolites that are relevant to disease or biology — than three other large databases, namely the Human Metabolome Database, MassBank, and the Lipid Metabolites and Pathways Strategy database, according to its developers.
In addition to containing more metabolites that are relevant to disease and biology, comparisons between the aforementioned repositories — using human serum and Escherichia coli samples — also showed that "there [are] a huge number of peaks … that we detect in biological sample[s] which are not in the database," indicating that there is "quite a lot of room for growth," Gary Patti, an assistant professor in Wash U’s chemistry department and a co-author on the Nature Biotech paper, told BioInform.
Scripps researchers began developing METLIN in 2004. A paper describing the resource that was published in Therapeutic Drug Monitoring the following year explained that the database was developed to address challenges with characterizing known and unknown metabolites by providing "a database of known endogenous metabolites, drugs, and drug metabolites and their respective structures; high-resolution FTMS data and MS/MS data from a number of standard biologic samples; and LC/MS proﬁles from a broad spectrum of samples and processing methods."
Over the last few years, METLIN's developers have partnered with a number of companies and research institutes, including Sigma-Aldrich, Cayman, and ChromaDex as well as labs at the Joint Bioenergy Institute and now Washington University to facilitate the acquisition of more molecules for which to generate MS/MS data. Today it contains more than 64,000 metabolites that are linked to outside resources such as the Kyoto Encyclopedia of Genes and Genomes for further reference and inquiry.
It's a far cry from the early days of metabolomics research where researchers might spend days combing databases and other sources of information to identify just one metabolite of interest, METLIN's developers said.
Previously, "if you were going to do a metabolomics experiment … one of the requirements [was] that you have a reference standard and if there [were] no database that [contained] a lot of reference standards, then you would end up having to purchase and run and analyze these standards yourself," Patti explained. METLIN "offers a huge repository of these reference standards," he said.
It's also a source for information about how metabolites fragment, Siuzdak added.
In proteomics, "you can theoretically generate fragment ions from the peptides that are being analyzed and those serve as the information for the identification of the protein, but with metabolomics, theoretical fragmentation is typically breaking all the bonds in the molecule and … [it's] not very meaningful because it actually doesn't account for rearrangements," he explained to BioInform. "As a result, having an actual experimental database of how these molecules fragment is incredibly important in terms of expediting the process of identification."
Efforts to increase the content of the database have also revealed a far more complex metabolome than previously thought, the developers said.
Rather than collect data from major metabolic processes alone, METLIN's creators incorporated compounds that aren't part of "canonical pathways" like the Krebs cycle, but which have turned out to be "important markers" of biology and disease progression in research studies, Patti said.
One project, done in Siuzdak's laboratory, used the database to identify a metabolite that is associated with chronic pain when found in higher than normal levels. They believe that finding ways to break down this metabolite or to prevent it from forming might lead to new treatment options for patients who suffer from the condition. The team is also working to identify metabolites involved in cancer progression that might offer targets for new therapies.
Meanwhile, Patti's lab at Wash U is investigating the role of metabolism in chronic pain and is also looking into metabolite roles in aging and drug addiction.
Other groups are using METLIN to explore metabolites produced after the ingestion of drugs as part of forensics work; as well as ways to detect signs of pesticide ingestion.