Nutrition & AI: Why We're Learning to Understand Food All Over Again

A visit to today's research laboratories quickly reveals why so many scientists speak of a genuine paradigm shift. For decades, nutrition science focused on around 150 well-known nutrients, including vitamins, minerals, amino acids, fatty acids, carbohydrates and a number of other bioactive compounds. Together, they continue to form the basis of almost every dietary recommendation. Today, however, a much bigger picture is emerging. Current databases describe more than 26,000 different chemical compounds that have already been identified in foods. Some researchers even believe that the true number may be significantly higher.

This does not diminish the value of what we already know. Vitamins remain vitamins. Protein remains protein. Dietary fibre has certainly not become less important. What has changed is the growing recognition that these familiar nutrients may represent only part of a much larger biological network. We are moving from a nutrient-centred view of food towards a systems-based perspective. Individual nutrients remain essential but increasing attention is now being paid to the way they interact with one another.

Only a few years ago, much of what scientists can now observe in modern laboratories would have remained completely invisible. Vitamins, fatty acids and amino acids could already be measured reliably, but they represented only a small fraction of the true molecular diversity found in food. Today, it is becoming increasingly clear that a food is less like a list of individual ingredients and far more like a finely tuned biological network.

The complete collection of currently known molecules within a food is referred to as its Foodome. You can think of it as the food's molecular landscape. A network of vitamins, fatty acids, proteins, polyphenols, aroma compounds and thousands of additional molecules. Every apple, every lentil and every blueberry possesses its own unique Foodome, extending far beyond the information shown in a traditional nutrition label.

Foodomics is the scientific discipline dedicated to studying this molecular landscape. The term combines the word food with the family of omics sciences. Including genomics, proteomics and metabolomics which all share a common principle: rather than analysing biological systems as isolated components, they seek to understand them as integrated, interconnected systems. Foodomics applies this systems-based approach to food and nutrition.

As a result, the central question of nutrition research is changing. For decades, researchers asked: Which nutrient produces which effect? Today, a much broader question is taking centre stage: What biological information does a food contain and how does the human body interpret it? AI is the tool. Foodomics is the scientific method. The real transformation, however, lies in our understanding of food itself. Food is no longer seen merely as a source of individual nutrients, but as a complex biological system whose components interact with one another and with the human body.

Modern mass spectrometers and other high-throughput technologies can now capture thousands of chemical signals within minutes. Powerful algorithms compare these datasets, identify similarities between molecular structures and model relationships on a scale that would be impossible for humans to analyse manually. This allows researchers to generate hypotheses about previously unknown molecules and their potential biological functions - hypotheses that can then be tested experimentally. Artificial intelligence is therefore not replacing scientific research. It is helping scientists analyse enormous volumes of data, uncover hidden patterns and develop entirely new research questions.

However, the Foodome of a food alone does not explain how it affects a particular individual. This is where Multiomics comes into play. This research approach combines different biological layers into a single, integrated perspective. The molecules found in food are analysed alongside a person's genes, gut microbiome, metabolic products and lifestyle factors such as sleep, physical activity, stress and hormonal changes.

This makes it easier to understand why the same meal can produce very different effects in two different people. Our genes influence how we process nutrients. The gut microbiome modifies components of our food before they even enter our metabolism. Physical activity, sleep and chronic stress all affect the very same biological pathways. As a result, nutrition, metabolism and lifestyle can no longer be viewed as separate disciplines. Yet even within this increasingly detailed map of food, a significant part remains unexplored. This has led to the emergence of the term Nutritional Dark Matter. It refers to the thousands of molecules found in our food whose existence can already be detected, but whose biological function, health relevance and interactions with other compounds are still largely unknown.

The comparison with astronomy is intentional. Astronomers believe that a large part of the universe consists of matter whose existence can be inferred, yet whose properties remain largely unknown. Our food appears to be remarkably similar. Researchers continue to identify thousands of measurable molecules whose biological significance is only beginning to be understood. These include rare plant compounds, only partially characterised polyphenols, sulphur-containing compounds, small peptides and molecules that are formed during cooking, fermentation or through interactions with our gut microbiome. Their potential role in metabolism, health, longevity and disease prevention is still largely unexplored.

Nutritional Dark Matter is therefore not the opposite of Foodome or Foodomics. Instead, it describes the still unknown part of the molecular world that Foodomics seeks to understand. It reminds us just how incomplete our understanding of food remains, despite decades of scientific progress. 

The true significance of artificial intelligence lies here. Not because it can provide better dietary advice than humans. Not because it replaces laboratory research or clinical studies. But because it makes visible what has remained hidden for so long: relationships. AI identifies patterns within a level of complexity that has long exceeded the capacity of any single scientific discipline. Individual molecules become interconnected networks. Isolated findings become meaningful relationships. And step by step, a collection of known and still unknown nutrients is evolving into an entirely new understanding of food. A blueberry is not a different fruit than it was ten years ago. An apple does not contain more molecules today than it did yesterday. What has changed is our perspective. And that may well be the true revolution.

Anyone following current research could easily conclude that, within a few years, every meal will require its own personal algorithm. Sensors will monitor metabolism in real time. Artificial intelligence will analyse blood biomarkers. The gut microbiome will be decoded. And digital systems will generate personalised nutrition plans. These scenarios may sound futuristic. Some of them will probably become reality. But the most exciting development has already begun. And it has less to do with technology than with a fundamentally new understanding of nutrition.

With every new study, we learn more about how differently people respond to the very same meal. Genes, the gut microbiome, age, sleep, exercise, hormonal changes and medication all influence the metabolic processes that take place after eating. Nutrition is therefore evolving from a one-size-fits-all recommendation towards a science that increasingly recognises the uniqueness of each individual. Artificial intelligence helps organise these enormous amounts of data and uncover relationships that previously remained invisible.

The implications extend far beyond personalised nutrition. Artificial intelligence is already transforming the way food is produced. In agriculture, intelligent systems help optimise water use, detect plant diseases at an earlier stage and predict crop yields with greater accuracy. Vertical farming is bringing food production closer to urban areas. Precision fermentation is opening up new possibilities for producing selected food ingredients. At the same time, cultivated meat has the potential to fundamentally change the way meat is produced. New production technologies may also contribute to using natural resources more efficiently. Taken together, these developments make one thing increasingly clear: In the future, nutrition will be understood as a connected system from soil and crops to food processing and, ultimately, the biological responses within the human body.

For the fitness and health industry, this shift in perspective opens up entirely new opportunities. Talking about nutrition will no longer be limited to calories, macronutrients or dietary supplements. The focus is expanding from individual nutrients to biological networks, from general recommendations to personalised responses, and from short-term effects to long-term health and longevity. Increasingly, the key question will be how food communicates with our metabolism, supports recovery, enhances performance and contributes to healthy ageing over the long term.

At the same time, scientific disciplines that have traditionally worked in isolation are beginning to converge. Nutrition science, food chemistry, microbiome research, systems biology, bioinformatics, medicine and artificial intelligence are evolving into a shared field of research. Step by step, the analysis of individual components is giving way to a systems-level understanding of nutrition. This is where the greatest potential for the years ahead lies.

With all these technological advances, one simple question naturally arises: What does this new knowledge mean for our everyday lives? Perhaps surprisingly, current research confirms many recommendations that have long been considered sound advice. A varied diet rich in vegetables, fruit, legumes, whole grains, nuts, seeds and fermented foods provides far more than the sum of its known nutrients. It exposes the human body to an extraordinary diversity of biological molecules many of which we are only just beginning to understand. The more closely scientists study food, the more convincing the value of dietary diversity becomes. It nourishes us not only with energy and essential nutrients, but also with a vast molecular diversity whose significance science is only beginning to uncover.

Perhaps this is the greatest surprise of all. For decades, scientific progress seemed to consist primarily of discovering ever smaller building blocks. Today, the perspective is expanding once again. Individual molecules remain important. But increasingly, it is their interaction that takes centre stage. Slowly, the individual pieces of the puzzle are forming a much bigger picture. And with that, we come full circle.

For decades, we described food largely in terms of what we were able to measure. Today, modern high-throughput technologies reveal a molecular diversity that reaches far beyond our previous models. Artificial intelligence then helps analyse these enormous datasets, identify patterns and model complex biological relationships. An apple is therefore not a different food than it was yesterday. What has changed is our understanding of it.

And that is precisely what makes scientific discovery so fascinating. The more our knowledge grows, the more clearly we recognise the limits of what we know. Scientific progress is not only about finding better answers. It is also about asking better questions.

Nutrition research has reached exactly this point. It builds upon decades of remarkable scientific discoveries while simultaneously uncovering an entirely new world that has, until now, remained hidden. Artificial intelligence opens the door. Science itself must walk through it. And perhaps that leads to the most exciting question of all: What will we know about food ten years from now that we cannot even imagine today?

Concept and Editorial: Concept and Editorial: Petra Orzech.

This article is based on current scientific publications in the fields of nutrition science, Foodomics, systems biology and artificial intelligence. A selection of the most important sources can be found below.

Sources and Further Reading

1.      Cifuentes A. Food analysis and foodomics. J Chromatogr A. 2009;1216(43):7109. doi:10.1016/j.chroma.2009.09.018. 
DOI: https://doi.org/10.1016/j.chroma.2009.09.018 
PubMed: https://pubmed.ncbi.nlm.nih.gov/19765718/

2.      Cifuentes A. Foodomics: Principles and Applications. In: Cifuentes A, editor. Foodomics: Advanced Mass Spectrometry in Modern Food Science and Nutrition. Hoboken (NJ): Wiley; 2013. p. 1–13. doi:10.1002/9781118537282.ch1. 
DOI: https://doi.org/10.1002/9781118537282.ch1 
Wiley: https://onlinelibrary.wiley.com/doi/10.1002/9781118537282.ch1

3.      Capozzi F, Bordoni A. Foodomics: a new comprehensive approach to food and nutrition. Genes Nutr. 2013;8(1):1–4. doi:10.1007/s12263-012-0310-x. 
DOI: https://doi.org/10.1007/s12263-012-0310-x 
Volltext (PMC): https://pmc.ncbi.nlm.nih.gov/articles/PMC3535000/

4.      Afendi FM, Ahnert SE, et al. Foodomics: A Data-Driven Approach to Revolutionize Nutrition and Sustainable Diets. Front Nutr. 2022;9:874312. doi:10.3389/fnut.2022.874312. 
DOI: https://doi.org/10.3389/fnut.2022.874312 
Volltext: https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2022.874312/full

5.      Shi X, et al. Foodomics: A sustainable approach for the specific nutrition and diets for human health. Food Chem Adv. 2024;5:100995. doi:10.1016/j.focha.2024.100995. 
DOI: https://doi.org/10.1016/j.focha.2024.100995 
Volltext (PMC): https://pmc.ncbi.nlm.nih.gov/articles/PMC11525469/

6.      Cifuentes A, Herrero M, Simó C, García-Cañas V. Foodomics: MS-based strategies in modern food science and nutrition. Mass Spectrom Rev. 2012;31(1):49–69. doi:10.1002/mas.20335. 
DOI: https://doi.org/10.1002/mas.20335 
PubMed: https://pubmed.ncbi.nlm.nih.gov/21374694/

For readers who would like to explore the topic in greater depth:

1.      PubMed – Foodomics

2.      Foodomics (Wiley Fachbuch)

3.      Foodomics – Open-Access-Review (Genes & Nutrition)

She is particularly fascinated by the latest discoveries at the intersection of nutrition science and artificial intelligence. As a member of a community-supported organic farm, she experiences first-hand every year just how much natural foods have to offer and how much we still have to learn about them. That curiosity became the inspiration for this article.