Metabolic networks represent genome-derived information about the biochemical reactions that cells are capable of performing. Mapping omic data onto these networks is important to refine model simulations. However, metabolomic data mapping remains very challenging due to difficulties in identifier reconciliation between annotation profiles and metabolic networks. MetaNetMap is a Python package designed to automatise the process of mapping metabolomic data onto metabolic networks. It includes several layers of identifier matching, the use of customisable databases, and molecular ontology integration to suggest the most matches between experimentally-identified metabolites and molecules defined in the network. We demonstrate its usability and the quality of automated mapping using two datasets. ### Competing Interest Statement The authors have declared no competing interest. Agence Nationale de la Recherche, ANR-22-PEAE-0011, ANR-11-INBS-0010

Today, metabolomics literature suffers from ambiguity in metabolites’ nomenclature, making difficult intercomparison between publications and can lead to misinterpretations. Progress in the implementation of FAIR principles in metabolomics in various scientific communities is therefore imperative for successful comparisons across studies and for moving towards more large-scale metabolomics. In this context, the aim of the present work was to explore the potential ambiguities that may be introduced during metabolite contextualization and reporting, and finally provide operational guidelines for metabolite name and identifier conversion to increase interoperability in metabolomics. From a list of 100 frequently annotated metabolites in human plasma, but also relevant for plant sciences, several workflows based on different existing identifier conversion tools were set up and evaluated, using two alternative approaches, one from an experimenter and the other from a data scientist's perspective. Findings showed a high level of mismatches using metabolite names as input, whereas starting from identifiers showed heterogeneity in the conversion consistency, depending on the association between input identifiers, algorithm of the selected tool, their respective versions, as well as versions of databases used for mapping. Errors in cross-reference databases were also highlighted. Despite these facts, InChIKeys were found to provide the highest quality results using all identifier conversion tools. From these results, operational guidelines were proposed using a curation process based on computational iterations, testing the stability and consistency of this conversion process, thus guaranteeing future metabolite contextualisation (e.g. links with pathways or phenotypes) and the interoperability of result reports.

Recent advances in light microscopy have transformed the scale and complexity of biological imaging data, creating an urgent need for sophisticated computational pipelines capable of extracting meaningful biological insights. However, the current software ecosystem for bio-image analysis remains highly fragmented, with researchers needing to integrate tools written in disparate programming languages and architectural paradigms. This fragmentation gives rise to 'dependency hell' and creates significant technical barriers for life scientists. We present the novel and flexible architecture of BioImageIT, a lightweight open-source workflow management system designed to bridge the gap between advanced computational tools and end-user bio-image analysts. Built upon Python, BioImageIT has evolved into a dual interface architecture: a node-based visual programming GUI alongside a comprehensive Python Application Programming Interface (API). The system features the Wetlands environment management system for automatic dependency resolution, and adopts pandas DataFrames as the universal data structure for inter-node communication. BioImageIT enforces adherence to FAIR principles (Findable, Accessible, Interoperable, Reusable) throughout the analysis lifecycle, automatically capturing comprehensive metadata for every processing step. The architecture abstracts the underlying computational infrastructure, laying the groundwork for seamless scaling from local workstations to high-performance computing (HPC) clusters-a capability currently under active development.

The genus Camellia comprises more than 200 evergreen species of major economic and ornamental importance, characterised by high morphological and chemical diversity. While several species have been extensively studied for their bioactive compounds, the metabolic basis of floral trait variation across the genus remains poorly understood. In this study, a predictive metabolomics framework was applied to investigate the relationship between leaf metabolic profiles and floral traits, focusing on flower colour and floral form. Leaves from 315 individual trees, including 15 Camellia species and representing 1,160 samples, were analysed by untargeted metabolomics, generating a large-scale metabolic profiling dataset. A dedicated quality control strategy was implemented to ensure analytical stability across multiple injection series and flowering seasons. Penalised generalised linear models were used to uncover robust metabolic predictors associated with floral traits and to evaluate model performance through internal and external validation. Distinct sets of metabolites were associated with flower colour and floral form, with limited overlap between traits. Predictive performance was consistently higher for colour than for floral form, indicating more structured metabolic signatures for chromatic traits. The selected predictors spanned multiple major chemical classes, supporting a systemic organisation of the metabolome rather than reliance on single biosynthetic pathways. Consistently high predictive accuracies were obtained, reaching approximately 87% for both flower colour and floral form, and remaining clearly above the corresponding no-information rates (≈ 43%). Together, these results demonstrate that leaf metabolomics can be used to robustly predict floral traits in Camellia and highlight the potential of predictive metabolomics as a tool for early phenotype inference, quality control and selection in long-lived ornamental species. ### Competing Interest Statement The authors have declared no competing interest.

Species-specific feedback between plants and soil microbial communities is an important driver of vegetation dynamics. Arbuscular mycorrhizal (AM) fungi colonise most terrestrial plants but are not expected to generate specific feedbacks due to low host specificity. We tested whether variation in mycorrhizal associations and associated rhizosphere metabolomes among co-existing temperate grassland species leads to species-specific plant-soil feedback. More mycorrhizal plant species showed more divergent plant-soil feedback: they experienced reduced growth and mycorrhizal colonisation in soils originating from weakly mycorrhizal species, but feedback became neutral in soil from species with similar mycorrhizal strategies. The species with the most self-promoting soil feedback was characterised by strong metabolome shifts related to stress and immune responses following soil inoculum manipulation, while the metabolomes of species with more negative feedback were unresponsive. This study demonstrates that AM fungi can generate species-specific plant-soil feedback, which can be predicted from plant mycorrhizal strategies and rhizosphere chemistry.

The genus Camellia comprises more than 200 evergreen species of major economic and ornamental importance, characterised by high morphological and chemical diversity. While several species have been extensively studied for their bioactive compounds, the metabolic basis of floral trait variation across the genus remains poorly understood. In this study, a predictive metabolomics framework was applied to investigate the relationship between leaf metabolic profiles and floral traits, focusing on flower colour and floral form. Leaves from 315 individual trees, including 15 Camellia species and representing 1,224 samples, were analysed by untargeted metabolomics, generating a large-scale metabolic profiling dataset. A dedicated quality control strategy was implemented to ensure analytical stability across multiple injection series and flowering seasons. Penalised generalised linear models were used to uncover robust metabolic predictors associated with floral traits and to evaluate model performance through internal and external validation. Distinct sets of metabolites were associated with flower colour and floral form, with limited overlap between traits. Predictive performance was consistently higher for colour than for floral form, indicating more structured metabolic signatures for chromatic traits. The selected predictors spanned multiple major chemical classes, supporting a systemic organisation of the metabolome rather than reliance on single biosynthetic pathways. Consistently high predictive accuracies were obtained, reaching approximately 87% for both flower colour and floral form, and remaining clearly above the corresponding no-information rates (≈ 43%). Together, these results demonstrate that leaf metabolomics can be used to robustly predict floral traits in Camellia and highlight the potential of predictive metabolomics as a tool for early phenotype inference, quality control and selection in long-lived ornamental species.

Metagenomics has lowered the barrier to microbial discovery–enabling the identification of novel microbes without isolation–but cultures remain imperative for the deep study of microbes. Cultivation and isolation of non-model microbes remains a major challenge, despite advances in high-throughput culturomic methods. The quantity of simultaneous experimental variables is constrained by time and resources, but the list can be reduced using computational biology. Given an annotated genome, metabolic modelling can be used to predict source nutrients required for the growth of a microbe, which acts as an initial screen to inform culture and isolation experiments. This chapter provides an overview of metabolic networks and modelling and how they can be used to predict the nutrient requirements of a microorganism, followed by a sample protocol using a toy metabolic network, which is then expanded to a genome-scale metabolic network application. These methods can be applied to any metabolic network of interest–which in turn can be created from any genome of interest–and are a starting point for experimental validation of source nutrients required for microorganisms that remain uncultivated to date.

The genus Camellia comprises more than 200 evergreen species of major economic and ornamental importance, characterised by high morphological and chemical diversity. While several species have been extensively studied for their bioactive compounds, the metabolic basis of floral trait variation across the genus remains poorly understood. In this study, a predictive metabolomics framework was applied to investigate the relationship between leaf metabolic profiles and floral traits, focusing on flower colour and floral form. Leaves from 315 individual trees, including 15 Camellia species and representing 1,160 samples, were analysed by untargeted metabolomics, generating a large-scale metabolic profiling dataset. A dedicated quality control strategy was implemented to ensure analytical stability across multiple injection series and flowering seasons. Penalised generalised linear models were used to uncover robust metabolic predictors associated with floral traits and to evaluate model performance through internal and external validation. Distinct sets of metabolites were associated with flower colour and floral form, with limited overlap between traits. Predictive performance was consistently higher for colour than for floral form, indicating more structured metabolic signatures for chromatic traits. The selected predictors spanned multiple major chemical classes, supporting a systemic organisation of the metabolome rather than reliance on single biosynthetic pathways. Consistently high predictive accuracies were obtained, reaching approximately 87% for both flower colour and floral form, and remaining clearly above the corresponding no-information rates (≈ 43%). Together, these results demonstrate that leaf metabolomics can be used to robustly predict floral traits in Camellia and highlight the potential of predictive metabolomics as a tool for early phenotype inference, quality control and selection in long-lived ornamental species.

Summary Plant ecological and evolutionary strategies are shaped by interactions between phylogenetic history and environmental constraints, resulting in leaf and stomatal traits. However, traditional trait-based and phylogenetic approaches often fail to fully explain biochemical mechanisms underlying ecological strategies, particularly for leaf and stomatal traits. Plant metabolomes integrate genetic, physiological, and environmental information and therefore represent a promising intermediate phenotype for investigating links between biochemical diversity, functional traits, and evolutionary patterns. We analysed metabolomic profiles from 74 plant species with various growth forms and ecological types. Using machine learning approaches, we explored whether metabolic variation could predict plant functional divisions, growth forms and phenological types, but also physiological traits related to drought resistance. Metabolomic data contained structured information associated with variation in plant functional traits, ecological strategies, and phylogenetic relationships. Machine learning models identified with high accuracy distinct metabolic signatures linked to differences among plant functional divisions, growth forms, phenology, and trait values. Our study demonstrates that predictive metabolomics provides a powerful and integrative framework to investigate plant ecological and evolutionary strategies. By linking biochemical diversity with plant phylogeny, and ecophysiological traits across multiple species, this approach offers new opportunities to explore the mechanistic basis of plant evolution.

Pearl millet is a high nutritional cereal recognised for its agro-climatic resilience, making it relevant for food security under climate change scenarios. Phenotypic traits are indicative of crop performance, stability and adaptability, yet the potential of metabolomics to predict these traits has not been explored. This study aimed to identify metabolite–trait associations in the Brazilian germplasm core collection, comprising 203 pearl millet genotypes, by combining untargeted metabolomics with machine-learning models. Grains metabolic profiles were obtained using untargeted UHPLC-LTQ-Orbitrap-HRMS. Phenotypic data were sourced from standardised evaluations conducted by Embrapa across different years and field trials within the Sete Lagoas experimental station (Minas Gerais, Brazil). Generalised linear modelling with penalisation (GLM) and Random Forest was applied to explore the correlation between metabolism and 21 phenotypic traits. GLM successfully predicted eight qualitative and seven quantitative traits. Prediction accuracy was higher for qualitative traits, reflecting their comparatively simpler genetic architecture, whereas quantitative traits also achieved satisfactory performance (R² ≥ 0.6). Key predictors included phenolic compounds, amino acids, fatty acids, and carbohydrates. Notably, several associations corresponded to metabolites involved in nitrogen metabolism and vegetative growth, underscoring biologically meaningful links between metabolic profiles and trait variation. This exploratory study presents the first metabolome characterisation of a pearl millet germplasm bank, coupled with predictive modelling of phenotypic traits. However, our findings are constrained by the single-environment design and the absence of population-structure assessment. To establish the stability and biological relevance of these results, future work should incorporate multi-environment trials and pathway-level analyses accounting for population structure.

Connecting the characterization of juvenile (pre-anthesis) plant stress responses in controlled environments to field agronomic performance is a challenge. The oilseed crop Camelina sativa (camelina), with its innate resilience and plasticity, presents an opportunity to understand the underlying mechanisms of juvenile resilience and identify the implications for yield in diverse pedoclimates. A better understanding of camelina's abiotic stress resilience is important in the context of climate change and the development of breeding programs for climate-tolerant crops. In this study, 54 accessions representing the genetic diversity observed in the wider publicly available population were used to investigate the plasticity of camelina's early stage response to drought and heat stress, combined with an evaluation of field performance in multilocation field trials. A combinatorial phenotyping approach of early stage drought and heat stress identified stress-responsive signatures within the diversity panel. The substantial variation in the morphophysiological line-specific responses to stress indicated that juvenile and mature camelina plants have significant plasticity and access different stress response strategies. In response to stress, we observed significant molecular metabolic adjustment alongside significant lipid remodeling and physiological compensation. Camelina was resilient to drought stress, and certain metabolites were identified as indicators of abiotic stress response. Applying an integrated approach, early stage phenotyping and multilocation field trials provided a complete assessment of the camelina stress response and facilitated a connection to crop productivity. This approach facilitates improved breeding programs, addresses the restrictions of limited genetic diversity in camelina, and supports the development of local varieties optimized for climate resilience.

Linking genotype and phenotype is a fundamental challenge in biology. In this respect, machine learning is playing a pivotal role in systems biology. As central phenotypic traits, fruit development and relative growth rate (RGR) result from interactions between gene regulation, metabolism, and environment. In the present study, we carried out a multispecies transcriptomic analysis of nine different fruits. To illustrate fruit transcriptomes, transcripts were first compared using multivariate methods, revealing similar main profiles. They were then used as variables to predict four growth traits, that is RGR, developmental progress, fruit weight, and protein content, using generalized linear models to decipher the mechanisms involving gene expression in development. The predictions were highly satisfactory despite disparities when the model did not include the entire panel of fruit species. Based on orthogroups derived from BLAST and annotated consensus sequences from gene ontology terminology, variables annotated for metabolic processes, especially those involving cell wall carbohydrates and proteins, were found to be the most effective in predicting growth. In addition, predictions were improved for RGR when introducing a 7 d lag between transcript contents and growth traits, suggesting the necessity of considering the proteins produced to enhance phenotypic trait predictions. These original results showed that growth traits can be predicted very well with generalized linear models based on orthogroups from multi-species transcriptomes.

Plants modulate their rhizochemistry, which affects soil bacterial communities and, ultimately, plant performance. Although our understanding of rhizochemistry is growing, knowledge of its responses to abiotic constraints is limited, especially in realistic ecological contexts. Here, we combined predictive metabolomics with soil metagenomics to investigate how rhizochemistry responded to environmental constraints and how it in turn shaped soil bacterial communities across stress gradients in the Atacama Desert. We found that rhizochemical adjustments predicted the environment (i.e. elevation, R2 between 96% and 74%) of two plant species, identifying rhizochemical markers for plant resilience to harsh edaphic conditions. These metabolites (e.g. glutamic and succinic acid, catechins) were consistent across years and could predict the elevation of two independent plant species, suggesting biochemical convergence. Next, convergent patterns in the dynamics of bacterial communities were also observed across the elevation gradient. Finally, rhizosphere predictors were associated with variation in composition and abundance of bacterial species. Biochemical markers and convergences as well as potential roles of associated predictive bacterial families reflected the requirements for plant life under extreme conditions. This included biological processes such as nitrogen and water starvation (e.g. glutamic and organic acids, Bradyrhizobiaceae), metal pollution (e.g. Caulobacteraceae) and plant development and defence (e.g. flavonoids, lipids, Chitinophagaceae). Overall, findings highlighted convergent patterns belowground, which represent exciting insights in the context of evolutionary biology, and may indicate unique metabolic sets also relevant for crop engineering and soil quality diagnostics. Besides, the results emphasise the need to integrate ecology with omics approaches to explore plant-soil interactions and better predict their responses to climate change.

This chapter explores advances and methodologies in high-throughput metabolic phenotyping through metabolomics and predictive modeling to enhance the understanding of plant metabolism. Key techniques, data analysis tools, and applications in plant science research are discussed. The potential of predictive modeling to identify new metabolic pathways and markers associated with plant performance and improve crop traits is highlighted. Future directions and challenges in the field are also examined.

Since the late 2010s, artificial intelligence (AI), encompassing machine learning and propelled by deep learning, has transformed life science research. It has become a crucial tool for advancing the computational analysis of biological processes, the discovery of natural products, and the study of ecosystem dynamics. This review explores how the rapid increase in high-throughput omics data acquisition has driven the need for AI-based analysis in life sciences, with a particular focus on plant sciences, animal sciences, and microbiology. We highlight the role of omics-based predictive analytics in systems biology and innovative AI-based analytical approaches for gaining deeper insights into complex biological systems. Finally, we discuss the importance of FAIR (findable, accessible, interoperable, reusable) principles for omics data, as well as the future challenges and opportunities presented by the increasing use of AI in life sciences.

In the context of climate change, temperature is a key abiotic driver of bunch microclimate, which, in order to reduce Botrytis cinerea development, is often managed in vineyards via practices such as leaf removal. The heat-dependent mechanisms of pathogen resistance in grapevines nevertheless remain to be fully elucidated. In this study, the effect of heat stress (HS) applied specifically to green bunches on infections caused by B. cinerea on ripe berries inoculated 23 days later was assessed for two years in a greenhouse. Bunches of the Cabernet Sauvignon (CS) and Merlot (M) cultivars were heated 6 days 8 h daily with a 10 °C increased temperature. In vitro bio tests highlighted a significant heat-enhanced resistance only in CS berries, whereas a stable constitutive resistance characterized the M berries. Bunch veraison and total sugar content were not affected by HS, rejecting its effect on maturation dynamics. Therefore, berry preformed barriers at the time of inoculation, which can hinder fungal colonization, were investigated. While HS had nearly no effect on waxes, it significantly affected the cutin content in both varieties, and more significantly its composition in CS. Similarly, the antifungal skin condensed tannins overaccumulated following HS in both cultivars, and their basal level was greater in CS than in M. Otherwise, M accumulated more stilbene and flavonoid compounds, which may have contributed to the observed varietal resistance. Finally, untargeted metabolomic data revealed a range of compounds modulated by HS in CS as potential candidates involved in resistance.

Mild winters are becoming increasingly common in temperate regions due to climate change, which may have important impacts on ecosystems and agriculture. In particular, rising temperatures affect the progression of winter dormancy—a crucial developmental stage in perennial plants—making tree development and reproduction particularly vulnerable to climate change. A better understanding of how future temperature conditions will disrupt dormancy in cultivated fruit trees is crucial for anticipating the impeding consequences and identifying potential adaptation strategies. We investigated the effect of very constrained temperature conditions, i.e. several levels of cold deprivation and early cold exposure, on sweet cherry flower buds during dormancy onset and maintenance, using phenological observations and transcriptomic analyses. We show that temperature is a major driver of dormancy progression as cold deprivation and early cold exposure strongly modify the timing of phenological phases as well as gene expression patterns. We identified genes and signaling pathways specifically activated and/or repressed by cold temperatures, and therefore potentially involved in the optimal progression of dormancy. Finally, thanks to an integrative analysis of molecular data obtained under natural and prolonged warm conditions, we characterize a distinct shallow dormancy phase induced by cold deprivation, with a unique molecular signature. Highlights Our phenological and molecular analysis of sweet cherry dormancy under constrained conditions reveals a gene expression timeline in response to temperature and uncovers a shallow dormancy stage induced by cold deprivation. ### Competing Interest Statement The authors have declared no competing interest.

IntroductionA better understanding of the physiological response of silage maize to a mild reduction in nitrogen (N) fertilization and the identification of predictive biochemical markers of N utilization efficiency could contribute to limit the detrimental effect of the overuse of N inputs.ObjectivesWe integrated phenotypic and biochemical data to interpret the physiology of maize in response to a mild reduction in N fertilization under agronomic conditions and identify predictive leaf metabolic and proteic markers that could be used to pilot and rationalize N fertilization.MethodsEco-physiological, developmental and yield-related traits were measured and complemented with metabolomic and proteomic approaches performed on young leaves of a core panel of 29 European genetically diverse dent hybrids cultivated in the field under non-limiting and reduced N fertilization conditions.ResultsMetabolome and proteome data were analyzed either individually or in an integrated manner together with eco-physiological, developmental, phenotypic and yield-related traits. They allowed to identify (i) common N-responsive metabolites and proteins that could be used as predictive markers to monitor N fertilization, (ii) silage maize hybrids that exhibit improved agronomic performance when N fertilization is reduced.ConclusionsAmong the N-responsive metabolites and proteins identified, a cytosolic NADP-dependent malic enzyme and four metabolite signatures stand out as promising markers that could be used for both breeding and agronomic purposes.