Agrifood systems refer to the interconnected set of processes, institutions, and infrastructures involved in the production, transformation, distribution, and consumption of food and agricultural products (FAO, 2021). Rather than functioning as simple linear supply chains, these systems are increasingly recognized as dynamic, multi-dimensional socio-ecological networks that influence livelihoods, nutrition, resource use, and innovation. Achieving sustainability in these systems depends on systemic innovation supported by enabling conditions such as regulatory stability, institutional alignment, social acceptance, and investment in research and infrastructure (Herrero et al., 2020). Consequently, agrifood systems are being redefined not only as food production, but as complex platforms for integrated, science-based transformation.

Recent developments in agrifood systems emphasize the integration of digital technologies to support strategic decision-making across increasingly complex value chains. The emergence of Agri-food 4.0 represents a major shift toward intelligent, data-driven frameworks designed to enhance efficiency, traceability, and environmental performance (Belaud et al., 2019). In this context, “omics” science has played a significant role in advancing agrifood research, with particular emphasis on metabolomics. Using advanced analytical instrumentation, high-throughput data acquisition, and machine learning techniques, metabolomics now stands at the forefront of modern food science (Taheri et al., 2024).

The term “Omics” refers to a set of integrative disciplines aimed at analyzing complex interactions within biological systems. It has been widely employed as a holistic framework for investigating biological processes through the application of scientific approaches such as genomics, metabolomics, proteomics, and transcriptomics (Misra et al., 2019; Dai and Shen, 2022). Among these, metabolomics has long been recognized as a relevant approach due to its ability to capture the biochemical phenotype of organisms in response to genetic and environmental factors (Fiehn, 2002). Through the identification and quantification of small-molecule metabolites such as amino acids, organic acids, sugars, and lipids, it provides a direct reflection of physiological and metabolic states. This makes metabolomics especially valuable for investigating responses to stress, disease, nutrition, and environmental changes across diverse biological systems. Within this context, the field of Foodomics has emerged as an integrative framework that brings together omics technologies, with metabolomics serving as a central tool for investigating food composition, nutritional function, and the biological mechanisms through which food components influence health (Cifuentes, 2009; Valdés et al., 2022).

Metabolomics stands out as an essential tool in agrifood systems, as it can be applied across different stages of food production and quality evaluation (Zhang J. et al., 2023). In plant-based systems, it supports cultivar characterization and helps elucidate metabolic processes involved in plant growth, fruit development, and ripening, providing information that can be used to improve product quality and extend shelf life. In food safety contexts, metabolomics is increasingly applied to postharvest studies, authenticity assessment, and traceability, as it enables the identification of metabolites associated with storage-related disorders, the evaluation of pesticide residues, and the detection of adulteration products. In addition, metabolomics is a highly relevant tool for identifying novel bioactive metabolites in food matrices, which may occur naturally in food species or be generated during food processing. Given its versatility and increasing accessibility, metabolomics is particularly valuable in regions where agriculture plays a strategic economic role. Its development and application are especially relevant in countries where crop and livestock production represent core pillars of the economy. This applies to Uruguay, which, with its strong agricultural foundation and prominent position in international food markets, holds significant potential to benefit from scientific advances in this field. The production and export of soybeans, meat, and citrus fruits are fundamental to the economy of the country, whereas olive oil and native fruits are promising new production activities.

This review explores the role of metabolomics within agrifood systems, highlights the strategic importance of agriculture in Uruguay, and analyzes scientific studies that apply metabolomic approaches to relevant agricultural products for Uruguayan economy. The primary objective of presenting this thematic overview is to emphasize the value of metabolomics as a scientific tool for advancing food production and supporting sustainable agricultural practices.

Metabolomics is an advanced analytical approach that enables the comprehensive profiling of low molecular weight metabolites in biological systems. In agrifood research, it plays a central role in linking genetic variation and environmental factors to phenotypes, supporting a deeper understanding of plant physiology, animal performance, and food quality (Fiehn, 2002; Tian et al., 2016; Anzano et al., 2022). For example, in soybean, metabolomic studies have shown that variation in amino acids and specialized metabolites, including isoflavones and other polyphenols, reflects the effects of genetic background, developmental stage, and abiotic stress, being associated with phenotypic traits related to seed development and stress responses (Cao et al., 2022). Moreover, in citrus, metabolomic studies indicate that variation in flavonoids, coumarins, terpenoids, sugars, organic acids, and volatile compounds is shaped by differences among species, fruit developmental stage, and environmental conditions, being associated with phenotypic traits linked to fruit maturation, stress responses, and sensory quality (Gaikwad et al., 2025).

Environmental factors are key drivers in metabolomic studies and can be broadly classified as abiotic and biotic. Abiotic factors refer to non-living stressors, including drought, salinity, temperature extremes, solar radiation, and nutrient limitations. In contrast, biotic factors involve living agents such as pathogenic microorganisms, insect pests, competitive weeds, and other biological interactions. By capturing the biochemical responses of organisms to these diverse challenges, metabolomics provides valuable insights for enhancing crop resilience, improving animal health, and optimizing system productivity. Its applications extend throughout the entire agrifood chain, from breeding and cultivation to processing, authentication, and nutritional evaluation.

Metabolomic approaches have become preferred tools for holistic investigations of metabolic processes. These studies can be performed using either untargeted strategies, which aim to maximize metabolite detection across chemically diverse compounds, or targeted strategies, which focus on the quantification of specific, preselected metabolites associated with defined chemical classes or metabolic pathways (Verpoorte et al., 2005). To achieve a comprehensive analysis of the metabolome, high-resolution and high-sensitivity analytical platforms are essential. The most employed techniques include Gas Chromatography-Mass Spectrometry (GC-MS), Liquid Chromatography–Mass Spectrometry (LC-MS), and Nuclear Magnetic Resonance (NMR) Spectroscopy (Verpoorte et al., 2008).

Nowadays, spectroscopic techniques are increasingly used in metabolomics due to their cost-effectiveness and rapid, non-destructive characteristics. In this context, Fourier-transform infrared (FTIR), near-infrared (NIR), and Raman spectroscopy provide fingerprint information reflecting the overall metabolic composition of agricultural products and can be more readily applied as routine analytical tools (Cebi et al., 2023). In contrast, metabolomic studies aiming at a deeper characterization of metabolic pathways and biomarker identification typically rely on more robust spectrometric platforms, such as GC-MS, LC-MS, and NMR, which allow detailed metabolite annotation and quantification (Wu et al., 2022).

One of the major challenges in metabolomic studies is dealing with complex biological matrices, which contain a high diversity of metabolites. In addition, sampling strategies, sample collection, and extraction procedures can significantly influence the resulting metabolic profiles. Furthermore, each analytical platform presents inherent limitations, as LC–MS is affected by ionization efficiency, GC–MS is restricted to volatile or derivatizable metabolites, and NMR spectroscopy has relatively low sensitivity in complex biological samples. However, advances in analytical technologies, including the integration of complementary analytical techniques and the continuous expansion of metabolomic databases, are progressively improving compound identification and enabling a more comprehensive characterization of complex metabolomes.

The interpretation of metabolomic data requires the use of advanced statistical and bioinformatics tools. Commonly applied approaches are unsupervised and supervised learning methods. The former is typically employed for exploratory data analysis and enables the identification of underlying patterns or groupings within the dataset, while the latter is widely used for biomarker discovery, classification, and predictive modeling (Ren et al., 2015). Accordingly, Principal Component Analysis (PCA) and Partial Least Squares-Discriminant Analysis (PLS-DA) are frequently applied as robust multivariate tools. Following these analytical steps, metabolomics helps uncover how metabolic pathways are organized and regulated, highlighting routes involved in natural product biosynthesis relevant to agricultural performance (Dixon et al., 2006). These insights support applications in breeding, agronomy, nutrition, and food innovation.

Beyond analytical challenges, data analysis in metabolomics also represents an important point of attention. Unsupervised multivariate analyses, such as PCA, may lead to overinterpretation of low-variance components, whereas supervised methods such as OPLS-DA are particularly susceptible to overfitting, especially in datasets with many variables and limited sample sizes. Machine learning has become an integral component of modern metabolomics, serving as a complementary or alternative strategy to classical linear models due to its ability to handle complex and high-dimensional data. Supervised methods, including random forests (RF), support vector machines (SVM), and artificial neural networks (ANNs), are increasingly used to improve predictive accuracy and facilitate the identification of biologically meaningful patterns. These approaches contribute to a more comprehensive understanding of metabolic regulation and system-level behavior (Liebal et al., 2020). To facilitate visualization, Figure 1 provides a schematic overview of the metabolomics workflow and its integration into agrifood systems. The process begins by identifying key challenges in agricultural production, such as abiotic and biotic stress, demands for quality, traceability, sustainability, and livestock productivity. Once samples are analyzed using platforms such as LC–MS, GC-MS, or NMR, the resulting data are processed and modeled. This enables the identification of biomarkers, which are metabolites associated with stress tolerance, growth regulation, or nutritional traits. In addition to biomarker discovery, metabolomics also supports the interpretation of metabolic pathways, helping clarify how metabolic processes respond to agricultural challenges. These findings can then be translated into practical improvements in agriculture, enhancing product quality, and contributing to a more sustainable production chain that delivers higher-value products to consumers.

The agricultural sector, together with livestock, fishing, and associated agro-industrial activities, plays a central role in Uruguay’s economy, contributing substantially to the country’s gross domestic product (GDP). In 2023, agrifood exports made up 80% of Uruguay’s total goods exported, underscoring the country’s reliance on agricultural trade (Uruguay XXI, 2024a). The main agricultural exports include soybeans, rice, wheat, citrus fruits, and olive oil, whereas livestock production remains the leading activity within Uruguay’s agrifood sector and a cornerstone of its export economy (Uruguay XXI, 2024b). Uruguayan meat is exported to more than 100 countries, with China, the United States, and the European Union among the most prominent destinations.

The National Institute of Agricultural Research (INIA) in Uruguay plays a central role in generating, adapting, and transferring technologies and knowledge that address the specific needs of the national agrifood sector (INIA, 2024). These institutional initiatives emphasize Uruguay’s growing capacity to incorporate advanced analytical tools, such as metabolomics, into agricultural research and innovation. Expanding the application of metabolomics across both crop and livestock systems could improve breeding strategies, enhance product differentiation, and optimize nutritional outcomes.

Currently, research groups in Uruguay are initiating metabolomic approaches to a range of nationally relevant products, including beef, soybean, olive oil and mandarins. Preliminary studies aimed to understand how factors like diet, genotype, and cultivation conditions influence nutritional composition, quality traits, and consumer acceptance. Examples include the impact of diet on beef quality (Luzardo et al., 2021), variability in pecan nutritional composition (Ferrari et al., 2022), sensory-driven selection of mandarin cultivars; (Migues et al., 2021; Migues et al., 2022), the effects of irrigation on olive oil quality (Conde-Innamorato et al., 2022), and support food safety of fruits and vegetables (Pereira et al., 2021).

Although metabolomics has been applied in some national studies, its full potential remains underexplored. Expanding its use could enhance agricultural productivity through the discovery of new biomarkers, optimization of protocols, and integration of advanced data analysis techniques. Investing in this field can support more precise decision-making, and drive innovations tailored to local crops, breeds, and environmental conditions. As the global demand for sustainable, high-quality agrifood products continues to rise, metabolomics represents a relevant path to enhance competitiveness within the agricultural sector.

Building on the strategic overview of Uruguay’s agricultural systems, this section reviews some of the recent international studies applying metabolomics to agrifood products of national relevance: soybean, meat, olive oil, citrus fruits, and native plant species. Together, these studies illustrate how metabolomics has been successfully applied worldwide to address similar production systems and commodities, reinforcing opportunities for scientific advancement at the national level.

The references used in this review were retrieved from the Scopus, PubMed and Google Scholar databases. Studies were identified using keywords such as metabolomics, metabolomic studies, and related variations, combined with terms referring to agrifood products of interest including soybean, meat, olive oil and citrus fruits. For the native plant species, fruit-bearing species of interest were considered: Butia odorata, Eugenia uniflora, Feijoa sellowiana, and Psidium cattleianum. The search results were filtered by publication year, and studies published more than 8 years prior were excluded. Following this, studies were screened based on analytical methodology, and only those employing NMR, LC-MS, or GC-MS approaches were included. The selected studies were then examined in greater detail, and those focusing on species or challenges not directly relevant to the Uruguayan agrifood context were excluded. Finally, for each agrifood product, the studies were grouped into thematic application domains, as summarized in Table 1. By synthesizing international evidence, this section provides a background that supports the expansion of metabolomics research within Uruguay’s agrifood sector.

Metabolomics has become an essential approach in agrifood research, increasingly applied to improve agricultural practices and ensure food security. By providing integrated insights into plant- and animal-based systems, it enables the identification of biomarkers associated with product quality, nutritional attributes, and environmental adaptation. These advances contribute to more informed decision-making in breeding, cultivation, livestock management, and postharvest processes, directly influencing productivity, product quality, and the sustainability of agrifood systems.

As previously mentioned, agrifood production plays a predominant position in Uruguay’s economy, led by commodities such as soybeans, meat, and citrus fruits, along with initial olive oil industry, and potential new crops derived from native fruits. In this review, we examined recent international studies that investigated these agrifood products using metabolomic approaches. Among the most frequently studied themes are: (i) metabolic features related to composition, nutritional value, quality, and functional attributes, and (ii) production and environmental responses, including cultivation practices, processing effects, traceability, and authenticity.

Within this context, metabolomics emerges as an effective strategy for developing high-quality agrifood products. Through the application of this approach and the use of appropriate analytical platforms, continued progress is expected, driving innovation in the development of resilient and sustainable agrifood systems.