Metabolomics in Personalized Nutrition

Metabolomics is a rapidly evolving field that involves the analysis of metabolites within biological systems; by capturing the metabolic profile of individuals, metabolomics offers insights into the underlying biochemical processes influenced by diet, genetics, and environmental factors, and enables the identification of metabolic pathways affected by diet, thereby facilitating the development of targeted dietary interventions. (Ulaszewska et al., 2019).  Metabolomics plays a crucial role in personalized nutrition by identifying metabolite biomarkers associated with specific dietary patterns, and with specific diseases and conditions (Miller et al., 2019).  For example, a study by Barfoot et al. (2021) found that consumption of wild blueberry was associated with both increased cognition and also an increase in a certain urinary metabolite. In addition, changes in metabolite concentrations over time could indicate progression toward a disease state (Gupta et al., 2023).  Despite its immense potential, metabolomics faces several challenges and limitations. These include the high dimensionality and complexity of metabolomics data, variability in sample preparation and analytical techniques, and the need for standardized protocols for data acquisition and analysis (Ulaszewska et al., 2019).

 

In the context of metabolomics, graph theory can be employed to model metabolic pathways, where metabolites are represented as nodes, and biochemical reactions as edges, which enables a deeper understanding of metabolic interactions and pathways through characterizations such as centrality, compactness, and assortativity, capturing the structural and functional relationships between metabolites and reactions, and facilitating the identification of key metabolic hubs and pathway dynamics (Klein et al., 2012).  Network analysis techniques such as path scoring allow for the identification of significant  metabolites and pathways; by assessing the importance and interconnectivity of nodes within metabolic networks, network analysis can highlight key metabolic features associated with specific physiological conditions or dietary interventions (Gill et al., 2022).  Visualization techniques aid in the interpretation and communication of complex metabolic networks, enabling researchers to explore and present their findings effectively (Ye et al., 2021).  Graph theory can also be used to model potential dietary modifications, representing the foods in a meal plan by the vertices with potential substitutions denoted by edges connecting the two; weights (a number) can be assigned to each edge depending on the desirability of the particular subsitution (Franco et al., 2019).  Michelini et al. (2022) used graph theory to evaluate the competition relationships among probiotic bacteria as part of their project to identify foods that could act as prebiotics for infants and help them develop a healthy microbiome.  They looked at the different metabolites from food that would in turn nourish the bacteria. 

 

Computational techniques such as linear programming optimization can be used to optimize meal planning to appeal to patient preferences and meet nutritional requirements (Sawal Hamid et al., 2019).  One strategy is to design individualized diets that meet nutritional requirements while remaining as close as possible to the diet each patient is used to by substituting nutrient dense foods for energy dense foods (Yu et al., 2018).  Salloum et al. (20220) describe meta-heuristic methods for generating meal plans, which could be incorporated into an automated interface.  They note that without an automated system like this, if people want to obtain personalized meal plans, they either need to have extensive knowledge about nutrition, the services of an expert nutritionist, or willingness to try an entire set of unfamiliar recipes.  They also describe case-based reasoning methods, which have not yet met with much success, and fuzzy reasoning methods, which basically just tell you whether the meal you are thinking of eating would be healthy or not.  Franco et al. (2019) note an advantage of fuzzy logic in that it is better suited to evaluate the words we typically use to describe food, which may not be easily reducible to the numbers used in most computational techniques. 

 

Machine learning techniques can be employed to develop predictive models for personalized nutrition (Chen et al., 2022).  Metaheuristic machine learning algorithms can be trained to generate meal plans for hospitals so that personalized nutritional recommendations for each patient that are based on individual characteristics can all be met while the kitchen only has to prepare 2-3 different meals (Ileri et sl., 2023). Martos-Barrachina  et al. (2022) used a metaheuristic to assemble meal plans meeting specific requirements from a database of recipes, with the idea that it could be expanded to allow people to specify their personal preferences and be provided with a meal plan that appealed to them personally.  Kirk et al. (2021) conducted a systematic review to see how machine learning is being used in the field of precision nutrition, where an individual’s specific responses to dietary factors are evaluated in order to provide dietary recommendations that take them into consideration.  They note that metabolomics is not yet incorporated to the extent that more traditional biomarkers are, but they anticipate it being used more commonly in the future.  They only found four papers looking at machine learning to make dietary recommendations, and none of them included metabolomics. As a challenge to precision in nutrition recommendations, they note that it is still difficult to accurately assess exactly what a person is eating on an ongoing basis. 

 

Personalized nutrition holds great promise in optimizing health outcomes by tailoring dietary recommendations based on individual characteristics such as variations in metabolite levels.  Graph theory and network analysis facilitate the exploration of metabolic pathways and interactions, while optimization techniques enable the optimization of nutrition plans. AI methods, including machine learning, leverage these foundations to generate personalized dietary recommendations. While challenges exist, such as data privacy, standardization of protocols, scalability, and ethical considerations, ongoing research and advancements in these areas will pave the way for more effective and precise personalized nutrition approaches. By leveraging the synergies between graph theory, optimization, and AI, personalized nutrition can transform dietary recommendations, improve health outcomes, and empower individuals to make informed choices for their well-being.

 

 

References:

 

Barfoot, K. L., Istas, G., Feliciano, R. P., Lamport, D. J., Riddell, P., Rodriguez-Mateos, A., & Williams, C. M. (2021). Effects of daily consumption of wild blueberry on cognition and urinary metabolites in school-aged children: a pilot study. European Journal of Nutrition, 60(8), 4263–4278. 

 

Bennet, D., Khorsandian, Y., Pelusi, J., Mirabella, A., Pirrotte, P., & Zenhausern, F. (2021). Molecular and physical technologies for monitoring fluid and electrolyte imbalance: A focus on cancer population. Clinical & Translational Medicine, 11(6), 1–23. 

 

Chen, S., Dai, Y., Ma, X., Peng, H., Wang, D., & Wang, Y. (2022). Personalized optimal nutrition lifestyle for self obesity management using metaalgorithms. Scientific Reports, 12(1). 

 

Das, R., Nag, S., & Banerjee, P. (2023). Electrochemical Nanosensors for Sensitization of Sweat Metabolites: From Concept Mapping to Personalized Health Monitoring. Molecules, 28(3), 1259. 

 

Franco, R. Z., Fallaize, R., Hwang, F., & Lovegrove, J. A. (2019). Strategies for online personalised nutrition advice employed in the development of the eNutri web app. Proceedings of the Nutrition Society, 78(3), 407–417. 

 

Gao, F., Liu, C., Zhang, L., Liu, T., Wang, Z., Song, Z., Cai, H., Fang, Z., Chen, J., Wang, J., Han, M., Wang, J., Lin, K., Wang, R., Li, M., Mei, Q., Ma, X., Liang, S., Gou, G., & Xue, N. (2023). Wearable and flexible electrochemical sensors for sweat analysis: a review. Microsystems & Nanoengineering, 9(1), 1–21. 

 

Gill, N. P., Balasubramanian, R., Bain, J. R., Muehlbauer, M. J., Lowe Jr., W. L., & Scholtens, D. M. (2022). Path-level interpretation of Gaussian graphical models using the pair-path subscore. BMC Bioinformatics, 23(1). 

 

Gonzalez, G., Gong, S., Laponogov, I., Bronstein, M., & Veselkov, K. (2021). Predicting anticancer hyperfoods with graph convolutional networks. Human Genomics, 15(1), 33. 

 

Gupta, S., Gormley, I. C., & Brennan, L. (2023). MetaboVariation: Exploring Individual Variation in Metabolite Levels. Metabolites (2218-1989), 13(2), 164. 

 

Ibrahim, N. F. A., Sabani, N., Johari, S., Manaf, A. A., Wahab, A. A., Zakaria, Z., & Noor, A. M. (2022). A Comprehensive Review of the Recent Developments in Wearable Sweat-Sensing Devices. Sensors (14248220), 22(19), 7670. 

 

Ileri, Y. Y., & Hacibeyoglu, M. (2019). Advancing competitive position in healthcare: a hybrid metaheuristic nutrition decision support system. International Journal of Machine Learning & Cybernetics, 10(6), 1385–1398. 

 

Kirk, D., Catal, C., & Tekinerdogan, B. (2021). Precision nutrition: A systematic literature review. Computers in Biology and Medicine, 133. 

 

Klein, C., Marino, A., Sagot, M.-F., Vieira Milreu, P., & Brilli, M. (2012). Structural and dynamical analysis of biological networks. Briefings in Functional Genomics, 11(6), 420–433. 

 

Liu, Y., Jiang, H., Qi, Y., & Yang, J. (2023). m-Health of Nutrition: Improving Nutrition Services with Smartphone and Machine Learning. Mobile Information Systems, 1–14. 

 

Martos-Barrachina, F., Delgado-Antequera, L., Hernández, M., & Caballero, R. (2022). An extensive search algorithm to find feasible healthy menus for humans. Operational Research: An International Journal, 22(5), 5231–5267. 

 

Michelini, S., Balakrishnan, B., Parolo, S., Matone, A., Mullaney, J. A., Young, W., Gasser, O., Wall, C., Priami, C., Lombardo, R., & Kussmann, M. (2018). A reverse metabolic approach to weaning: in silico identification of immune-beneficial infant gut bacteria, mining their metabolism for prebiotic feeds and sourcing these feeds in the natural product space. Microbiome, 6(1). 

 

Miller, I. J., Peters, S. R., Overmyer, K. A., Paulson, B. R., Westphall, M. S., & Coon, J. J. (2019). Real-time health monitoring through urine metabolomics. NPJ Digital Medicine, 2(1), N.PAG. 

 

Moore, J. B. (2020). From personalised nutrition to precision medicine: the rise of consumer genomics and digital health. Proceedings of the Nutrition Society, 79(3), 300–310. 

 

Neves, P. A., Simões, J., Costa, R., Pimenta, L., Gonçalves, N. J., Albuquerque, C., Cunha, C., Zdravevski, E., Lameski, P., Garcia, N. M., & Pires, I. M. (2022). Thought on Food: A Systematic Review of Current Approaches and Challenges for Food Intake Detection. Sensors (Basel, Switzerland), 22(17). 

 

Pedersen, H., Diaz, L. J., Clemmensen, K. K. B., Jensen, M. M., Jørgensen, M. E., Finlayson, G., Quist, J. S., Vistisen, D., & Færch, K. (2022). Predicting Food Intake from Food Reward and Biometric Responses to Food Cues in Adults with Normal Weight Using Machine Learning. Journal of Nutrition, 152(6), 1574–1581. 

 

Pérez, D., & Orozco, J. (2022). Wearable electrochemical biosensors to measure biomarkers with complex blood-to-sweat partition such as proteins and hormones. Microchimica Acta, 189(3), 1–28. 

 

Rein, M., Ben-Yacov, O., Godneva, A., Shilo, S., Zmora, N., Kolobkov, D., Cohen-Dolev, N., Wolf, B.-C., Kosower, N., Lotan-Pompan, M., Weinberger, A., Halpern, Z., Zelber-Sagi, S., Elinav, E., & Segal, E. (2022). Effects of personalized diets by prediction of glycemic responses on glycemic control and metabolic health in newly diagnosed T2DM: a randomized dietary intervention pilot trial. BMC Medicine, 20(1). 

 

Russo, S., & Bonassi, S. (2022). Prospects and Pitfalls of Machine Learning in Nutritional Epidemiology. Nutrients, 14(9), 1705. 

 

Salloum, G., & Tekli, J. (2022). Automated and personalized meal plan generation and relevance scoring using a multi-factor adaptation of the transportation problem. Soft Computing: A Fusion of Foundations, Methodologies and Applications, 26(5), 2561–2585. 

 

Sánchez-Tirado, E., Agüí, L., González-Cortés, A., Campuzano, S., Yáñez-Sedeño, P., & Pingarrón, J. M. (2023). Electrochemical (Bio)Sensing Devices for Human-Microbiome-Related Biomarkers. Sensors (14248220), 23(2), 837. 

 

Sawal Hamid, Z. B., Rajikan, R., Elias, S. M., & Jamil, N. A. (2019). Utilization of a Diet Optimization Model in Ensuring Adequate Intake among Pregnant Women in Selangor, Malaysia. International Journal of Environmental Research & Public Health, 16(23), 4720. 

 

Shinn, L. M., & Holscher, H. D. (2021). Personalized Nutrition and Multiomics Analyses: A Guide for Nutritionists. Nutrition Today, 56(6), 270–278. 

 

Sim, D., Brothers, M. C., Slocik, J. M., Islam, A. E., Maruyama, B., Grigsby, C. C., Naik, R. R., & Kim, S. S. (2022). Biomarkers and Detection Platforms for Human Health and Performance Monitoring: A Review. Advanced Science, 9(7), 1–29. 

 

Ulaszewska, M. M., Weinert, C. H., Trimigno, A., Portmann, R., Andres Lacueva, C., Badertscher, R., Brennan, L., Brunius, C., Bub, A., Capozzi, F., Cialiè Rosso, M., Cordero, C. E., Daniel, H., Durand, S., Egert, B., Ferrario, P. G., Feskens, E. J. M., Franceschi, P., Garcia-Aloy, M., … Vergères, G. (2019). Nutrimetabolomics: An Integrative Action for Metabolomic Analyses in Human Nutritional Studies. Molecular Nutrition & Food Research, 63(1), e1800384. 

 

Ye, G., Zhang, J., Bi, Z., Zhang, W., Zhang, M., Zhang, Q., Wang, M., & Chen, J. (2021). Dominant factors of the phosphorus regulatory network differ under various dietary phosphate loads in healthy individuals. Renal Failure, 43(1), 1076–1086. 

 

Yu, K., Xue, Y., Zhao, W., Zhao, A., Li, W., Zhang, Y., & Wang, P. (2018). Translation of nutrient recommendations into personalized optimal diets for Chinese urban lactating women by linear programming models. BMC Pregnancy and Childbirth, 18(1). 

 

Yuan, X., Li, C., Yin, X., Yang, Y., Ji, B., Niu, Y., & Ren, L. (2023). Epidermal Wearable Biosensors for Monitoring Biomarkers of Chronic Disease in Sweat. Biosensors (2079-6374), 13(3), 313. 

 

Zeevi, D., Korem, T., Zmora, N., Israeli, D., Rothschild, D., Weinberger, A., Ben-Yacov, O., Lador, D., Avnit-Sagi, T., Lotan-Pompan, M., Suez, J., Mahdi, J. A., Matot, E., Malka, G., Kosower, N., Rein, M., Zilberman-Schapira, G., Dohnalová, L., Pevsner-Fischer, M., … Segal, E. (2015). Personalized Nutrition by Prediction of Glycemic Responses. Cell, 163(5), 1079–1094.