The metabolomics profile, the complete inventory of a biological system’s metabolites, roils with activity, the ups and downs of its many constituents. And this activity, which reflects normal physiological processes or disease states, may alert investigators to metabolites of interest—potential biomarkers of disease, usually. Less commonly, but with increasing frequency, metabolomics studies are identifying regulatory molecules.

By capturing the dynamism of biological processes, metabolomics profiles complement genomics profiles, which are relatively static. For example, a metabolomics profile may reveal, in real time, which disease risks predicted by a genomics profile are being realized.

Such insights, however, have a price. They require that investigators overcome unique workflow and data processing challenges. Whereas a typical clinical laboratory study may focus on a handful of small molecules, a metabolomics study must keep track of hundreds or thousands of molecules.

Clearly, the metabolomics field, like other “omics” fields, needs to be comprehensive. Metabolomics, however, faces additional difficulties. For example, it relies on mass spectrometry and other separation methods that may not fully differentiate all small molecules in the sample. To acquire adequate discriminatory power, investigators may need to perform further experiments, which can complicate workflows and strain data processing resources.

In the metabolomics field, technologies are being developed to enable the efficient capture and analysis of large volumes of metabolomics data. Four such technologies are highlighted in this article. Encouragingly, each of these technologies is leading to advances in the study of human disease.

Weighty distinctions

Difficulties characterizing large numbers of metabolites for individual patients have limited the application of metabolomics in precision medicine and drug discovery. To overcome these difficulties, investigators are turning to artificial intelligence (AI). AI can accomplish characterization tasks that would otherwise require decades of research and billions of dollars.

AI is at the heart of a metabolomics technology developed by ReviveMed. The technology, which consists of a proprietary database and machine learning algorithm, not only identifies metabolites, it also integrates the resulting metabolomic datasets with other large-scale molecular datasets for use in developing new therapeutics. This technology is so promising that in 2018, it helped ReviveMed win Frost and Sullivan’s North American AI-Driven Next-Generation Metabolomics Technology Innovation Award.

figure shows an overview of ReviveMed’s technology
The figure shows an overview of ReviveMed’s technology, which combines mass spectrometry instrumentation and an artificial intelligence platform. The technology can identify metabolite peaks that arise due to dysregulated molecular mechanisms, pointing to potential drug targets.

“Even though metabolites provide the most functional information about diseases, their data has been underutilized,” says Leila Pirhaji, PhD, ReviveMed’s founder and CEO. That is, data about metabolites can be ambiguous because metabolites themselves can be hard to distinguish experimentally. Metabolites display heterogeneous structures, show rapid turnover, and present other obstacles. Fully characterizing metabolites requires multiple experimental procedures. For example, mass spectrometry can separate metabolites by weight, but different molecules may have the same weight, so further analysis is needed to distinguish them.

That’s where AI comes in. By developing an AI platform to identify metabolites based on mass, ReviveMed is bridging the gap in typical metabolomics studies. When metabolites are identified, they can be used to discover molecular mechanisms behind disease. This can point to potential drug therapies that have never been explored. ReviveMed is focusing on nonalcoholic fatty liver disease as one of its first indications.

ReviveMed’s technology is a network-based machine learning algorithm for integrative analysis of untargeted metabolomics data with other data such as genomics and proteomics data.

Nonalcoholic fatty liver disease is an ideal disease to study with a metabolomics approach. The disease, which occurs when excess fat accumulates in liver cells, is metabolic in nature. It is untreatable at present, and it is the number one reason for liver transplants in the United States.

“We have a unique way of looking at more data points than any other platform to understand the novel biology of the disease,” asserts Pirhaji. ReviveMed’s goal is to partner with pharmaceutical companies for clinical development of its programs. The company plans to look at 50 other indications in the next five years.

Biomarkers of frailty

A new collaboration between Metabolon, the Canadian Frailty Network, the Canadian Longitudinal Study of Aging (CLSA), and the McMaster Institute for Research on Aging highlights a unique application of metabolomics. The project will focus on biomarkers of frailty in aging to determine the severity of frailty and what can be done to help avoid the condition. It will use samples from Canada’s largest comprehensive study on aging to generate metabolomic profiles and identify biomarkers.

The CLSA is a research platform focusing on aging in 50,000 Canadians. Its goal is to identify the biological mechanisms of frailty, which previously have not been well understood.

Greg Michelotti, PhD
Greg Michelotti, PhD, Scientific Director, Metabolon

Metabolon’s scientific director, Greg Michelotti, PhD, says the initial analysis will focus on baseline samples from 10,000 participants from the CLSA analyzed for metabolomic and inflammatory biomarkers linked to frailty. Detailed lifestyle and clinical information will be combined with metabolomic analyses to uncover disease mechanisms and deepen researchers’ understanding of disease progression across a wide range of conditions.

Metabolon uses liquid chromatography and tandem mass spectrometry (LC/MS/MS) to perform metabolic profiling of small molecules in the samples. The company’s Precision Metabolomics™ platform has produced results including biomarkers, diagnostic tests, and mechanistic insights into disease in approximately 10,000 completed projects.

“The beauty of our approach is that it’s unbiased,” declares Michelotti. “We’re not going in with any presuppositions or hypotheses.” The study will screen 1200–1300 molecules from up to 70 biological pathways in an untargeted manner.

The idea is that by having knowledge of metabolic changes with aging in a very large group of study participants, it will be possible to characterize a healthy metabolome, and to pinpoint how individuals may deviate from the healthy reference range.

“That’s why metabolomics is so powerful,” explains Michelotti. “A lot of work is focused on genetics, but that really only identifies risk of disease. Metabolomics is unique in that it integrates genetic and nongenetic information.” That nongenetic information includes lifestyle, diet, and the microbiome. Consequently, it may yield a more comprehensive picture of health status.

Lipid movements

Nonalcoholic steatohepatitis (NASH), an advanced form of nonalcoholic fatty liver disease that can progress to cirrhosis and liver failure, is increasingly prevalent. Still, there are no approved treatments for the disease.

Biopharmaceutical companies are increasingly interested in investigating therapies for NASH. A new partnership between Barc Lab, a major international central laboratory, and Owl Metabolomics, a small company focusing on NASH and other liver diseases, will provide clinical support for those research programs.

Owl’s technology can identify more than 1000 small molecule metabolites in a single sample. This yields a comprehensive snapshot of the metabolome that can be used to study the metabolic changes underlying NASH. Owl refers to its technology as next-generation precision metabolomics. It uses a liquid chromatography and mass spectrometry–based platform tailored for liver biomarkers.

The company’s approach to NASH has been based on tracking the movement of lipids in the liver. “We realized a long time ago,” says Pablo Ortiz, MD, PhD, Owl’s CEO, “that if we could really track the movement of the lipids, we would have a clue to find a noninvasive diagnostic product for NASH.”

Owl developed two noninvasive assays for fatty liver screening and NASH diagnosis. The OWLiver Care test is a diagnostic test for nonalcoholic fatty liver that has been validated in comparison to liver biopsy. The OWLiver test is an early, noninvasive blood test that can identify people at risk of developing NASH, allowing them an opportunity to try lifestyle changes to head off the disease.

“This kind of synergy is very provocative,” Ortiz emphasizes. “We are creating a new tool to impact development.” This new collaboration will leverage Owl’s technology and Barc Lab’s analytical capabilities to support biopharmaceutical discovery programs searching for treatments for NASH.

Google map of the cell

When large amounts of metabolomics data are being generated, it’s not always immediately clear how the scientific researcher will leverage that data to deliver an actionable result. A co-marketing agreement between Sciex, a life sciences analytical technology company, and Elucidata, an integrated omics platform company, will allow researchers to process metabolomics data from a broad range of workflows supporting applications such as target identification and characterization of metabolic pathways. EluciData’s omics platform, Polly, will be promoted in conjunction with Sciex’s Triple TOF, X500R QTOF, and QTRAP systems, as well as Sciex’s SelexION differential mobility device.

On the Polly platform, workflows integrate omics data related to cellular phenotypes. Workflows available on Polly include Polly Metscape, a global metabolomics profiling data analysis application, and PollyPhi, a hypothesis validation tool that allows scientists to go from pathways to highlighting changes in enzyme function by analyzing the flow of labels through metabolites.

“If we’re going to enable precision medicine, we first need to enable the researchers who are at the forefront of this data to biology challenge,” says Baljit Ubhi, PhD, market manager, metabolomics and lipidomics, Sciex. “There’s a breaking point in metabolomics and even further downstream with other omics, where users can generate data at very high speeds. That’s where they’re stuck.”

“How,” she asks, “are they going to take the thousands of samples they’ve generated data on and do something meaningful with the data?” According to Ubhi, the reporting of omics data can become a critical bottleneck in a project. Through Elucidata’s collaboration with Sciex, project timelines can be reduced from months to weeks or even less.

“The way we see it, our mandate is to integrate different kinds of data that can help scientists reveal the underlying biology of a phenotype—a Google map of the cell, if you will,” says Abhishek Jha, PhD, co-founder, Elucidata. “Beyond that, we allow any scientific researcher to push through their datasets and map that data to the biology, allowing them to make decisions more efficiently.”

Metabolomics researchers typically use open source tools that need to be customized and configured for their projects. Elucidata believes that combining the Polly platform with Sciex metabolomics workflows will enhance the user experience of researchers, freeing them to spend more time on high-value discoveries.

Previous articleTaking Cell Lines to the Bank
Next articleCRISPR Vehicles Break Down Barriers to In Vivo Genome Editing