This article is based on an interview with Farhana Pinu and also on Pinu, Goldansaz, and Jaine’s “Translational Metabolomics: Current Challenges and Future Opportunities” from the Metabolites journal.
Metabolomics is a hot research area right now, promising a deeper understanding of organisms and better, more personalized diagnostics and treatment.
Researchers are making great progress every year, but so far, only a few metabolomics applications have made it all the way to the clinic.
Farhana and her colleagues gathered recommendations on how more studies can lead to successful clinical and industrial products. This is the science of translational metabolomics: “Translating” scientific discoveries into everyday products.
Aligning academic and industry goals
To understand why it’s challenging to translate research breakthroughs into real-world products, we need to understand a fundamental difference between science and industry. Science and industry have very different goals.
- Discovery drives science: the goal is to understand how things work.
- Products drive industry: the goal is to package research into easy-to-use, efficient, and profitable tools.
Science and industry often work on different time frames because of these different goals. Scientists might spend years or decades on the same problem, whereas the time frames for clinical applications last from several months to a few years.
Translational metabolomics aims to align these goals, and it has already shown some major successes. The most popular fields, which are most likely to be profitable, are:
- Prenatal tests assessing the risk of unborn babies contracting diseases;
- Hereditary cancer tests finding out if patients are at higher risk of developing certain cancers; and
- Personalized medicine tailoring treatment based on numerous factors, besides a patient’s illness.
The existing real-world successes in these areas make it likely that many more are on the way, but this does not mean that it is easy.
Challenges in translational metabolomics
The promise of translational metabolomics is clear, but plenty of obstacles remain in its way. Here are the main challenges and their potential solutions.
CHALLENGE #1: The translation process can fail at multiple stages
For an experiment to succeed and create concrete value, it needs:
- 🧪 robust experimental design;
- 🚚 sound data acquisition;
- 📊 easy-to-use data mining;
- 🤓 in-depth interpretation; and
- ✅ successful validation of candidate biomarkers.
Failing any of these steps can cause an entire project to break down, so quality assurance is paramount at each stage.
SOLUTION #1: Stringent processes and user-friendly tools
More stringent processes will help ensure the validity and soundness of experiments. This includes:
- Using international guidance: if researchers follow guidelines from professional and regulatory bodies more closely, research can be shared globally and achieve more success. Extensive documentation can seem overwhelming to scientists, but the breakthroughs in applied research pay off in the end.
- Developing user-friendly tools: if scientists can access user-friendly and open-source software and databases, especially for web-based data analysis, they will find global collaboration more efficient and reliable.
Better processes and tools will help ensure results are reliable enough to be reproduced by other researchers.
🤔 CHALLENGE #2: Negative perceptions
Metabolomics is still a niche field, not a household name. As with the other omics fields, it is sometimes seen as “hyped”, a field that over promises and under delivers.
Metabolomics is ambitious and does make big promises. In time, it’s supposed to give us pieces of the puzzle that will cure diabetes, cancer, and many other diseases. Instant diagnoses and all kinds of medical tooling are also meant to improve our lives. But so far, the world-changing tools haven’t materialized.
Metabolomics has been successful. But big wins have earned little publicity because they have been more niche and localized: the field needs better PR.
SOLUTION #2: Publicity and engagement from researchers
Farhana Pinu points out that scientists don’t always publicize the big successes of metabolomics well enough. They need to make more promotional efforts through social media and other channels.
The need to highlight results is common to other omics fields, too. But metabolomics is one of the less mature omics fields and is playing “catch-up” in some ways. The huge mass of data the field generates, besides it’s many goals, mean its successes can get buried.
The metabolomics community needs to acknowledge the perception problem and actively publicize its impact.
🤑 CHALLENGE #3: Metabolomics is an expensive game
Researchers use expensive analytics software and platforms that often seem over complicated. A full metabolomics lab costs millions of dollars to set up. An LC-MS platform alone costs nearly a million dollars.
To save money, laboratories often use small sample sizes, and this hurts results.
SOLUTION #3: Less expensive, miniaturized equipment
Anyone interested in metabolite measurement should be able to undertake a study. But without cheaper lab equipment, only a select few can access the equipment they need.
Developing miniaturized instruments that are cheaper and more accessible will act as a multiplier for other researchers. Vendors that produce hardware and software need to collaborate to make this happen.
🧑🔬 CHALLENGE 4: Multidisciplinary expertise
In other disciplines, research teams are generally made up of experts from the same field. But for translational metabolomics, teams need biologists, analytical chemists, statisticians, data scientists, and bioinformaticians.
The sheer variety of expertise means metabolomics teams are larger and more difficult to coordinate. This makes these teams more expensive and less efficient, which adds to the challenge of establishing multidisciplinary teams.
SOLUTION #4: Multidisciplinary collaboration and interdisciplinary training
Academic fields tend to specialize more over time. For translational metabolomics to succeed, we need to encourage both,
- Better collaboration between experts; and
- Upskilling and training of multidisciplinary experts.
A researcher experienced in biology and data science can often achieve results faster than a pair of experts representing both fields but working together without understanding each other’s expertise. But it’s increasingly difficult for one person to understand enough about a broad field like metabolomics.
For better collaboration, we need more spaces for experts to collaborate. Forums and symposiums can facilitate cross talk between fields and also between academic, industry, and regulatory bodies.
For multidisciplinary experts, we need specialized training programmes to teach individuals the unique interdisciplinary skill set required for metabolomics.
📊 CHALLENGE 5: Acquiring data accurately
Even with expensive equipment, there’s no universal extraction method for all metabolites. Many metabolites are unstable and can degrade from light, heat, and oxidation, besides other things, even when stored at -80 degrees.
Some metabolites have high concentrations and some have low concentrations. Each of these require different specialized protocols and equipment to accurately measure them.
Acquiring a complete metabolome of a single biological entity remains challenging, expensive, and time consuming.
SOLUTION #5: Better equipment and analytics platforms
Equipment and analytics software are often difficult to use. We need better hardware and software to acquire and analyze data accurately.
Once researchers have ensured they have good quality data, they also need easier ways to share that data globally, and between different omics fields.
Better data analysis and data sharing platforms will help translational efforts enormously.
🔢 CHALLENGE #6: Absolute quantification in metabolomics
Researchers often focus on “semi-quantitative” data. That is, their research is conclusive in their own lab and on their own equipment, but the hard numbers result from their specific setup and equipment.
For example, a researcher might find a metabolite that identifies a disease. They can then discriminate successfully between healthy and sick patients by looking at the quantity of this metabolite.
But this is a relative measure that isn’t calibrated to global standards. A doctor on the other side of the world can’t use these results on their own patients because there is no absolute threshold of how much of that metabolite sick people have compared to healthy people.
SOLUTION #6: Better calibration methods and global data-sharing
Working with absolute quantification is more challenging but far more valuable. This involves using carefully calibrated machines to identify exact quantities of specific metabolites that can be replicated by other labs around the world.
The effort is costly, in both equipment-calibration and quality assurance. But the value gained by sharing results globally more than compensates.
The Metabolomics Standards Initiative (MSI) monitors and reviews data and standards globally, and researchers need to follow these standards wherever possible to make it easier to turn experiments into clinical applications.
💅 CHALLENGE #7: Researchers are drawn to “glamour” topics
Everyone wants to "cure cancer" or "fix diabetes". These are big, ambitious projects that probably won't translate into real products any time soon. Most of them are unfeasible from an industry perspective: unlikely to be solved in a reasonable time frame and unlikely to be profitable.
Smaller and more localized problems are often more practical and more profitable. But they attract much less media attention so are often less attractive to funders and researchers, too.
While ambitious research is important, it’s more important to solve specific, validated problems faced in clinics.
SOLUTION #7: Simple and concrete outcomes
Researchers should partner with industry to ensure experiments have appropriate scope and context. This means experiments should be:
- Achievable in a reasonable time frame; and
- Targeted at an adequately-sized (even if niche) need.
If researchers at the hypothesis stage focus on results that can be translated into real-world applications (for example, portable devices), it is more likely their results will be used by industry.
🛠️ CHALLENGE #8: Inadequate end-user equipment
We know that research equipment needs to be smaller and cheaper. We also need advances in equipment for end users, such as doctors.
Some end-user products that follow translated studies can be difficult or time-intensive to use. The results might be hard to understand or require specialized analysis, leading to poor adoption.
SOLUTION #8: Easy-to-use equipment
Just as with lab equipment, we need simple end-user devices as well. Devices should:
- Be easy to use;
- Be portable;
- Produce rapid results;
- Produce easy-to-interpret results;
- Be inexpensive; and
- Be approved for use.
These priorities often lie beyond academic and scientific research, but translational metabolomics needs to consider them from the start.
Do you have metabolomics research to translate?
Our team has extensive experience using machine learning to convert metabolomics research into real-world solutions. If you’re working on metabolomics-related problems, we’d love to hear from you.