Plant Communications

Groundnut (Arachis hypogaea) is an important oilseed crop, and its production is severely affected by drought worldwide. Improving the drought-tolerance of groundnut without compromising its agronomic traits is a major challenge for researchers. Here, a meta-transcriptome analysis identified tandemly duplicated G{beta} subunits (AhG{beta}1 and AhG{beta}3) as early signaling genes under water deficit conditions. Heterotrimeric G-protein {beta}-subunit is a major signaling element that can regulate diverse responses in plants. Heterologous complementation studies in Arabidopsis G{beta} mutant (agb1-2) demonstrated evolutionary conservation of AhG{beta} function, while the transgenic overexpression and RNAi knockdown in groundnut established AhG{beta}1 as a positive regulator of plant growth, yield, and drought tolerance. Integrated transcriptomic, transcription factor regulatory network, and metabolomic analyses demonstrated that AhG{beta}1 coordinates drought-responsive transcriptional and metabolic reprogramming characterized by suppression of cell wall-associated processes and activation of phenylpropanoid, flavonoid, and isoflavonoid biosynthesis. This reprogramming resulted in enhanced accumulation of antioxidant flavonoids, including quercetin and rutin, reduced ROS accumulation, and improved drought tolerance. Furthermore, exogenous application of these flavonoids enhanced drought recovery, highlighting their role as protective downstream metabolites. Collectively, our findings establish AhG{beta}1 as a conserved regulator linking early G-protein signaling with transcriptional and metabolic reprogramming to maintain redox homeostasis during drought stress and identify promising molecular and metabolic targets for improving drought resilience in groundnut.

Iron (Fe) homeostasis is fundamental to plant growth and development and is centrally regulated by a well-conserved network of basic helix-loop-helix (bHLH) transcription factors. In Arabidopsis thaliana, bHLH121 regulates the expression of FER-like iron deficiency-induced transcription factor (FIT) and clade Ib bHLH38/39/100/101 genes, among others. However, its functional conservation remains unexplored, and some contradictory results have been reported regarding the effects of bHLH121 overexpression. Here, we investigated the evolutionary conservation of bHLH121 through complementary tests in the Arabidopsis bhlh121 mutant created by CRISPR/Cas9 gene editing. We also investigated the effects of ectopically expressing full-length, N-terminal, and C-terminal fragments of bHLH121 on plant responses to Fe deficiency in Arabidopsis. Transgenic plants overexpressing the full-length bHLH121 driven by the cauliflower mosaic virus 35S promoter (bHLH121 ox) resulted in shorter roots and more severe leaf chlorosis compared to wild-type plants under Fe limiting conditions. Moreover, overexpression of the N-terminal fragment (N121 ox) rendered plants highly sensitive to Fe deficiency, whereas overexpression of the C terminal fragment (C121 ox) did not. Consistent with the phenotypic observations, N121 ox plants accumulated lower Fe content, and the ferric chelate reductase activity was significantly reduced. In N121 ox plants, the induction of key Fe regulatory genes, especially FIT, was significantly inhibited under Fe deficiency. Confocal imaging showed that both N121-GFP and C121-GFP fusion proteins could localize to the nucleus. Electrophoretic mobility shift assay showed that both full-length bHLH121 and N121 could effectively bind the probes derived from the promoter of FIT. The present work reveals a conserved function of bHLH121 in Fe homeostasis across several dicot species and demonstrates that the N121 may possess a dominant nature in Fe deficiency responses.

Although iron (Fe) is abundant in soils, its low solubility, particularly under alkaline conditions, often limits plant Fe acquisition. Under these conditions, coumarin-mediated Fe mobilization becomes increasingly important. Arabidopsis thaliana FE DEFICIENCY-INDUCED KELCH REPEAT PROTEIN 1 (FDK1, AT3G07720) is strongly induced by Fe deficiency and regulated by the central transcription factor FIT, but its function is unknown. We investigated FDK1 using genetic, physiological, metabolite, and transcriptomic analyses. Loss of FDK1 altered root growth, whereas this phenotype was restored in the FDK1-Ox overexpression line. Under alkaline Fe deficiency, growth medium from FDK1-Ox contained strongly increased levels of the Fe-mobilizing coumarins fraxetin and esculetin compared with wild type, whereas their induction was attenuated in fdk1-1. RNA-seq analysis linked FDK1 abundance to coumarin biosynthesis and iron-homeostasis gene expression. FIT and many FIT-associated genes were highly expressed in FDK1-Ox, whereas several upstream iron-homeostasis regulators, including bHLH Ib factors, were highest in fdk1-1. These findings identify FDK1 as a component of the alkaline Fe-deficiency response that promotes iron-mobilizing coumarin accumulation and is associated with coordinated regulation of iron-homeostasis genes.

Amaranthus tricolor is a climate-resilient orphan crop whose betalain pigments contribute to nutritional quality and stress tolerance. However, breeding efforts have traditionally relied on field yield trials due to the limited availability of genomic information and genetic resources. To address this gap, we first used the Illumina and PacBio sequencing platforms to assemble a novel reference genome of A. tricolor, using the Japanese cultivar Bayam. By combining this reference genome with a core collection developed in our previous study, we aimed to identify genomic regions associated with pigmentation-related traits across winter and summer seasonal environments. Betalains are antioxidant phytonutrients that contribute to abiotic stress tolerance, making the regulation of their biosynthesis under heat stress particularly relevant for developing climate-resilient crops such as A. tricolor. As a result, ATR1.0CH03_5131103, which is associated with ATR1.0ch03g000565, was identified under the winter seasonal environment. Meanwhile, transcriptomic profiles in this study revealed that ATR1.0ch03g000565 was differentially expressed under both control and heat stress conditions, suggesting that it may be associated with the regulation of CYP76AD1 expression under optimal conditions. Under the elevated-temperature treatment, transcript levels of CYP76AD1 and 5GT decreased in Am087, coinciding with lower betacyanin-dominated absorbance indices. cDOPA-like homologs were upregulated; however, pigmentation-related absorbance remained reduced. Further molecular experiments, such as CRISPR-Cas9 knockouts or overexpression studies, are necessary to validate the regulatory mechanisms of CYP76AD1 and 5GT under heat stress.

Common tansy (Tanacetum vulgare), a significant agricultural invasive species, threatens forage productivity, livestock carrying capacity, and plant biodiversity across North American agroecosystems. Despite being a close relative to the ornamental Asteraceae Chrysanthemum, T. vulgare genomic studies are severely lacking, but remain critical to the development of novel weed management tools and the investigation of plant invasion evolution. In this study, we performed the assembly and annotation of the diploid genome (2n=2x=18) of an invasive North American T. vulgare genotype using long-read PacBio HiFi and Iso-Seq sequencing technology. The assembled haplotype genome comprised 9 pseudomolecules with a total size of ~4.8 Gb and a scaffold N50 of 546 Mb. In addition, both the plastid and mitochondrial genome sequences were assembled. BUSCO analysis revealed a high overall genome completeness with 69,343 predicted protein-coding genes (AED<0.6: 98%). A high-quality North American invasive reference T. vulgare genome, complementing a recent one assembled for a native European accession, is critical for the investigation of plant invasion biology, genome evolution, and the future development of novel genetic biocontrol weed management technologies.

Chlorophyll fluorescence imaging is a powerful tool used to characterize abiotic and biotic stress response of photosynthesis in plants. However, extracting metrics such as Fv/Fm (maximum quantum yield of photosystem II) from chlorophyll fluorescence images remains a substantial bottleneck in large measurement campaigns, often requiring significant manual effort. To resolve this issue, we developed fvfmPy, a user-friendly program which enables rapid analysis of chlorophyll fluorescence images of arrays of leaf discs. fvfmPy uses algorithmic image segmentation to identify regions of interest and extract fluorescence metrics from tens or hundreds of leaf samples per image with little to no manual input. fvfmPy is developed in Python and includes a graphical user interface enabling intuitive parameter adjustment and real-time quality control checking. We additionally developed fvfmR, an R wrapper package for fvfmPy. Fv/Fm values extracted with fvfmPy were indistinguishable from those obtained using manual methods (r2 > 0.99) but required 94% less time to complete (13 s vs. 200 s per 100 measurements). fvfmPy and fvfmR were built to directly integrate with existing fluorescence software packages and can greatly increase throughput of fluorescence measurements to help address key challenges in agriculture and ecology.

Palladium (Pd), a technology critical platinum group metal, is increasingly lost to the environment through anthropogenic activity, yet the molecular basis of Pd uptake and tolerance in plants remains poorly understood. Here, we examine the role of the Arabidopsis thaliana (Arabidopsis) COPT2 copper transporter in Pd transport and plant stress responses. Expression of COPT2 in Saccharomyces cerevisiae conferred Pd sensitivity, indicating a capability for Pd transport. Arabidopsis copt2 loss-of-function mutants accumulated less Pd in shoots and exhibited reduced reactive oxygen species (ROS) production, supporting a direct role for COPT2 in Pd distribution and associated oxidative stress. Our subsequent transcriptome profiling revealed extensive changes under Pd exposure, including downregulation of HMA2 and aquaporins, and upregulation of HMA7, glutathione transferases, and glutamine synthetase GLN1;1, consistent with the activation of metal detoxification and redox homeostasis pathways. Together, these findings identify COPT2 as a key transport protein mediating the uptake of Pd in Arabidopsis and suggest that Pd induces similar detoxification mechanisms as zinc (Zn), copper (Cu) and cadmium (Cd), providing a foundation for engineering plants for precious metal recovery.

Drought stress significantly reduces tomato (Solanum lycopersicum L.) productivity, and early detection is critical to minimize yield losses through timely interventions. In this study, we developed Drought-Spec-Net, a hybrid 1D convolutional neural network that integrates local and global spectral feature extraction to detect early drought stress from visible and near infrared (Vis NIR) spectra data of tomato seedlings. The model was trained on 378 samples using an 80:20 train test split, with 20% of the training set reserved for validation. DroughtSpecNet outperformed the evaluated baseline and state of the art models, achieving 97% accuracy, 95% precision, 98% recall, and an F1 score of 97%. To improve the agronomic interpretation of the model outputs, predicted drought probabilities were converted into a literature-informed potential yield impact indicator using a maximum impact level of 60%. On the test set (76 samples), mapped potential yield-impact values ranged from 0% to 60%, with an average reduction of 12.97%. We also conducted an initial experiment using our greenhouse RGB dataset, collected daily from drought treated and well-watered tomato plants at West Virginia State University (WVSU). From this dataset, 44 images were selected for ilastik-based canopy segmentation, producing plant-level drought severity indices (DSI) with a mean of 0.28, median of 0.14, and range of 0.01 to 0.91. Additionally, we trained and fine-tuned a large language model (LLM) based on PLLaMA7BInstruct, called AgriLLaMA, for automated agronomic report generation from Drought-Spec-Net outputs. The generated reports summarize predicted stress levels, mapped potential yield impacts, and preliminary management considerations. This integrated approach not only improves early drought stress detection but also delivers quantitative and interpretable estimates of potential productivity losses, providing a complete framework connecting physiological stress detection to actionable agricultural outcomes.

Photosynthesis is a key determinant of crop productivity, yet conventional gas-exchange measurements are labor-intensive and unsuitable for high-throughput assessment of dynamic photosynthetic acclimation across large genotype panels. Consequently, genetic variation in photosynthetic acclimation remains poorly characterized and difficult to exploit in breeding programs. Here we present a computational pipeline that integrates rapid optical sensing (chlorophyll meters and hyperspectral reflectance) with a mechanistic model of photosynthetic protein-turnover to estimate dynamic nitrogen allocation among photosynthetic components. The pipeline was demonstrated using 60 winter wheat (Triticum aestivum L.) cultivars sampled at 10 time points spanning leaf emergence to senescence. Uncertainties associated with each pipeline component were quantified and benchmarked against gas-exchange ground-truth measurements under three controlled-environment light and temperature regimes. Dynamic nitrogen allocation to light harvesting, electron transport, and carboxylation was characterized using three biologically interpretable parameters: maximum synthesis rate, degradation rate, and age-dependent decline in synthesis. Although hyperspectral models showed moderate predictive accuracy for photosynthetic capacity, prediction errors were predominantly random, allowing robust parameter estimation when observations across the leaf lifespan were integrated. The pipeline successfully resolved genotype-by-environment interactions in photosynthetic acclimation, with environmental and interaction effects contributing more strongly than genotype main effects to nitrogen dynamics. These results demonstrate a scalable, uncertainty-aware sensor-to-trait architecture for dynamic physiological phenotyping that enables high-throughput germplasm screening across diverse environments and provides novel physiological selection targets for improving crop adaptation under variable climates.

Abstract: Picrotoxane-type sesquiterpene alkaloids, including dendrobine, are characteristic constituents of Dendrobium nobile, but the enzymes that produce their proposed skeletal precursors remain in completely characterized. Here, we combined tissue metabolite profiling with genome-guided screening and biochemical characterization of terpene synthases (TPSs). GC-MS analysis revealed variation in sesquiterpenoid profiles and dendrobine signals among tissues and developmental stages. Manual curation of the chromosome-level genome yielded 16 full-length TPS-a candidates. Functional screening in an FPP-supplying Escherichia coli system showed that three closely related genes on chromosome 1, DnTPS1~DnTPS3, produced copaborneol as their dominant product. The principal product of DnTPS1 was purified and identified by NMR spectroscopy, and a purified-enzyme assay confirmed its direct conversion of FPP to copaborneol. Fluorescence imaging in rice protoplasts and Nicotiana benthamiana leaves indicated predominantly cytosolic localization of DnTPS1. Guided by sequence comparisons and structural modeling, we tested 12 DnTPS1 variants and found that substitutions at W377 and I295 altered the relative distribution of sesquiterpene products. Phylogenetic and regional gene-order analyses placed DnTPS1~DnTPS3 in the same clade and identified related TPS candidates within corresponding genomic regions of six Dendrobium species. These findings establish that three D. nobile TPSs can form copaborneol and identify DnTPS1 residues that influence product selectivity, providing a basis for testing the proposed connection to dendrobine biosynthesis.

Efficient yield prediction under field conditions remains a key challenge for crop breeders and agronomists. YieldLearn is an interpretable machine-learning pipeline developed to estimate yield-related traits in durum wheat (Triticum durum) using physiological indicators measured by the Hansatech Instruments Pocket PEA. The workflow integrates fluorescence parameters (Area, Fo, Fm, Fv/Fm, PI Inst.) with field-measured agronomic traits (Booting, Height, Maturity, Lodging, Yield kg/plot, and Adjusted Yield kg/ha). Five supervised learning algorithms multinomial logistic regression, random forest, support vector machine (radial kernel), k-nearest neighbors, and decision tree were evaluated using five-fold cross-validation. Models achieved average classification accuracy around 0.57 with AUC values exceeding 0.58. SHAP (SHapley Additive exPlanations) visualizations revealed that high Fv/Fm and PI Inst. values were positively associated with yield potential, whereas elevated Fo indicated stress-induced yield reduction. An interactive Shiny App enables real-time prediction and feature-importance exploration. Given the small sample size (165 plots) and absence of environmental covariates such as temperature or soil moisture, YieldLearn should be regarded as a reproducible exploratory framework rather than a production-ready predictor. Nonetheless, it offers a valuable foundation for guiding experimental design, validating physiological indicators, and supporting the next generation of scalable, interpretable crop-yield models.

Agrobacterium-mediated transformation is a cornerstone of plant genetic engineering, yet many crops remain recalcitrant to transformation and its efficiency is highly species- and genotype-dependent. The limited diversity of disarmed agrobacterial strains available for biotechnology, and the technically challenging workflows typically required for strain modification, constrain efforts to develop improved plant transformation platforms. Here, we describe the development of AGROtools-a versatile, expandable, and user-friendly modular cloning plasmid collection comprising more than 75 individual components, that is designed to simplify genetic engineering of agrobacteria. Constructs assembled with this toolkit couple visual reporters with selectable and counterselectable markers to facilitate identification of recombinant clones, thereby reducing reliance on PCR-based screening. We constructed and tested reporter cassettes that are functional across diverse agrobacterial strains and genomic contexts. By linking a visual reporter combinatorially with various antibiotic resistance markers and orthogonal site-specific recombinase recognition sequences, AGROtools can enable marker removal and recycling for up to four successive genomic modifications. We demonstrate the utility of AGROtools by disarming 10 novel wild-type strains, and accomplishing serial modifications within individual strain backgrounds. In addition to conventional recA-mediated homologous recombination, the collection incorporates components for applying contemporary strain engineering approaches, including CRISPR/dCas9 base editing, INTEGRATE-mediated targeted insertion, and phage-derived recombineering systems. Constructs can be assembled and deployed rapidly, enabling isolation of strains carrying an intended modification in less than two weeks. AGROtools provides a broadly applicable resource for accelerating strain development and genetic studies in agrobacteria, with potential to advance plant biotechnology by aiding improvement of disarmed strains.

The Fusarium oxysporum species complex (FOSC) comprises numerous plant pathogens infecting crops such as tomato, lettuce, melon, and banana. Despite its economic and ecological importance and decades of research, the species names linked to different phylogenetic lineages within this species complex remain largely controversial. Earlier taxonomic frameworks relied on a limited number of marker genes, variable taxon sampling and inconsistent species definitions, leading to a fragmented and unstable taxonomy. To address these challenges, we here analysed nearly 2,500 conserved marker genes from 660 genomes, including genome assemblies from 22 ex-type strains that were sequenced using long-read sequencing technology, representing nearly half of the 51 currently described species within FOSC. Maximum likelihood- and coalescence-based analyses of the full phylogenomic data matrix as well as subsets yielded a robust phylogeny of FOSC, which was furthermore corroborated by whole genome comparisons. Many previously described species are redundant, while two novel lineages remain undescribed. Direct comparisons to traditional marker gene-based approaches demonstrate their inadequacy for species delimitation in FOSC. We show that our phylogenomic framework can guide the detection of novel marker genes able to distinguish the major lineages within FOSC, which will be essential for taxon delineation and classification in future studies. Thus, our phylogenomic analyses provide a stable cornerstone for revising FOSC taxonomy and offers a critical resource for the broader research community.

DNA methylation dynamics regulate diverse plant stress responses, but the mechanisms governing arsenic (As) tolerance through epigenetic pathways remain largely unknown. Here we report that Arabidopsis mutants lacking the DNA demethylase, REPRESSOR OF SILENCING (ROS1), show substantially enhanced As resistance. Integrating transcriptome and methylome data, we found that PIN-FORMED 2 (PIN2), encoding an auxin efflux carrier that also transports As out of cells, represents a key ROS1 target. In ros1, we observed increased methylation in the last exon of PIN2, a region functioning as the promoter for an antisense long non-coding RNA (lncRNA). Hypermethylation of the promoter suppresses the antisense lncRNA, which follows reduction of sense PIN2 transcription. As a result, PIN2 expression increases in ros1 mutants during As exposure, enhancing As tolerance. Analysis of ros1 pin2 double mutants confirmed that PIN2 acts downstream of ROS1 in As tolerance. Our study demonstrates an epigenetic pathway in which ROS1 regulates As tolerance by controlling PIN2 expression through antisense lncRNA-mediated transcriptional interference, offering new epigenetic strategies using plants in heavy metal contaminated environments.

Lipolytic enzymes play critical roles in plant development, growth, and stress responses. In oat (Avena sativa), they are of particular interest because lipid hydrolysis contributes to rancidity and reduced grain quality, yet little is known about genomic variation in oat lipolytic enzymes. Here, we investigated genomic and expression diversity of the GDSL-esterase/lipase gene family using pangenome, transcriptome, and genome-wide association analyses. Across 22 A. sativa and two A. sterilis genome assemblies, we identified 7,775 GDSL genes (~324 per assembly), revealing wide conservation of ancestral orthologs alongside extensive copy-number variation among homoeologs and paralogs. Transcriptome profiling across 20 genotypes and six tissues demonstrated preferential expression in embryo and root, with tissue-specific patterns highlighting candidates for specific functions, such as seed lipid metabolism. Associations with structural variants further uncovered candidate mutations underlying expression variation among accessions. Comparisons of cultivated and wild oat detected loci with signatures of breeding-based selection, suggesting that GDSL genes have contributed to diversification of the cultivated gene pool. Using 629 whole-genome sequenced accessions, we also identified associations between GDSL regions and agronomic traits, particularly plant height and effective tiller number. Several regions showed patterns consistent with selection, highlighting candidate loci for further investigation of agronomic variation. Together, our results reveal substantial genomic and transcriptional diversity within the oat GDSL gene family and identify candidate genes and genomic regions for improving seed quality and agronomic performance. More broadly, our study demonstrates how pangenomic resources can be used to investigate gene-family diversity and function in complex polyploid crops.

The Unfolded Protein Response (UPR) maintains endoplasmic reticulum (ER) homeostasis during proteotoxic stress while also contributing to plant immunity. However, the functional relationship between UPR signaling, pathogen-associated molecular pattern (PAMP)-triggered immunity (PTI) and salicylic acid (SA) remains poorly understood. Here, we investigated this interplay using Arabidopsis mutants impaired in the two major UPR branches. Mutants defective in the INOSITOL-REQUIRING ENZYME 1 (IRE1)-bZIP60 branch displayed enhanced resistance to Pseudomonas syringae pv. tomato DC3000 and to the type III secretion-deficient strain hrcC-, indicating potentiation of PTI. Consistently, the ire1a ire1b double mutant exhibited increased flg22-induced callose deposition, whereas reactive oxygen species production was unaffected. Despite enhanced basal resistance, all UPR mutants retained full responsiveness to benzothiadiazole, demonstrating that downstream SA signaling remains functional independently of UPR integrity. Conversely, acute ER stress induced by tunicamycin promoted SA accumulation and extensive leaf necrosis. Reverse genetics revealed that SA-deficient mutants were largely protected from ER stress-induced cell death, whereas mutants over-accumulating SA exhibited enhanced susceptibility, demonstrating that endogenous SA is required for the full execution of this response. Analysis of an independent transcriptomic dataset further showed that loss of IRE1 de-represses genes involved in SA biosynthesis, PTI and defense under ER stress. Together, our results identify IRE1 as a key regulator coordinating ER homeostasis with immune signaling by limiting SA-dependent defense activation during ER stress, while simultaneously preventing excessive cell death.

Root nodule symbiosis allows legumes to meet their nitrogen requirements. However, the causes of symbiosis failure at high temperatures remain unknown. We demonstrated that elevated temperatures affect both the host and the microbial partner in Medicago truncatula and chickpea. Rhizobial viability and nod gene expression persisted under heat stress, but plant transcriptome analysis revealed a strong downregulation of the nodule organogenesis machinery. Nodule organogenesis was restored by constitutive overexpression of NIN-LBD16-NFYA1-NFYB16, miR2111, and CLE5 knockdown but failed to restore nitrogen fixation. In contrast, the genetic manipulation of the competence pathway involving CEP peptides significantly improved both nodulation and nitrogen fixation. Constitutive loss of the autoregulation of nodulation (AON) pathway in the SUNN alleles also rescued root nodulation and nitrogen fixation. Therefore, heat stress perturbs the balance between AON and the competence pathway, and genetic manipulation of these pathways is a promising approach to engineer thermotolerant nitrogen fixation in crop legumes.

Genetically encoded biosensors are central to plant cell biology, but low reporter signal can limit their use. Here, we developed modular intronised fluorophores to improve reporter output in transient and stable plant expression systems. Matched nuclear-localised triple mCitrine and mScarlet3 reporters were compared with and without introduced introns in Nicotiana benthamiana and Arabidopsis thaliana. Intronisation increased fluorescence, with the magnitude of enhancement depending on the fluorophore and expression context. Immunoblot analyses supported increased reporter accumulation. In Arabidopsis, intronised reporters driven by the tissue-specific SCARECROW promoter, produced stronger fluorescence signals that remained detectable farther from the root tip than non-intronised versions. Intronised versions of the abscisic acid (ABA) biosensor ABACUS2 retained ABA responsiveness and showed a lower frequency of reduced fluorescence between generations. These modular components provide a practical toolbox for improving fluorescent reporter detection and maintaining usable biosensor expression, supporting their application in plant cell biology and quantitative imaging.

Non-destructive methods for real-time phenotyping of root growth and development are essential to accelerate the breeding of climate-resilient crops. We present DIRTNet (Digital Investigations of Root Traits Net), a hybrid deep learning model that interprets in-soil signals collected by Fiber Bragg Grating (FBG) sensors. We first evaluated DIRTNet in a controlled simulation study in which metal rods of varying diameters were inserted into soil to emulate root growth. Using augmented training data with an 80:20 train-test split, DIRTNet surpassed baseline models, achieving 0.97 accuracy for root depth prediction and 0.94 for root diameter prediction. We then applied the best-performing models to maize (Zea mays) roots monitored by six FBG sensors placed at different soil location and depths. A control trial and a drought-stress trial yielded 192 samples, split 80:20 into training and holdout sets; the hold-out portion was divided equally into validation and test sets. Augmentation was applied only to the training data. On real maize root data, DIRTNet achieved 0.92 accuracy for both root depth and root width prediction across variable growth periods, covering root mass lengths up to 20 cm and diameters up to 10 cm. The prediction accuracy is independent of the soil moisture and drought condition. In a fixed-interval analysis, accuracy reached 0.97 for days 25-42 after planting and 0.93 for days 25-48, demonstrating robust performance as root complexity increased.

Reproductive stage drought disrupts flag leaf metabolism in rice (Oryza sativa), limiting photoassimilate translocation to developing grains and reducing grain yield (GY). How natural variation in this response relates to GY remains poorly understood. This study combines untargeted metabolomics of flag leaves from 271 indica rice accessions grown under well-watered (WW) and reproductive-stage water-limited (WL) field conditions with genome-wide association studies of 89 annotated polar metabolites. A total of 180 metabolite quantitative trait loci (mQTLs) were identified: 93 detected only under WL, 68 only under WW, and 19 under both conditions, indicating a predominant stress-dependent genetic control; mapping of GY itself detected only four marginal QTLs, none overlapping the mQTLs. Four high-density regions (HDRs) of mQTLs were identified: HDR1-1, detected only under WW conditions and associated mainly with raffinose family oligosaccharides; HDR1-2, HDR4, and HDR11, detected only under WL conditions and associated mainly with tricarboxylic acid cycle intermediates and amino acids. After correcting for genetic relatedness, accessions carrying the favourable haplotype at both HDR1-2 and HDR4 yielded 21.5% more under WL conditions than accessions carrying the unfavourable haplotype at both loci and lost 11.4 percentage points less GY to stress. Accessions carrying the favourable haplotype at only one locus showed a comparable reduction in yield loss. Haplotype-guided variant detection against the 3,000 Rice Genomes Project panel identified major-impact variants in candidate genes, including a polyphenol oxidase at HDR1-2 and tandem cytochrome P450s at HDR4. These hubs are promising, metabolically defined targets for breeding drought-resilient rice.