Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Deep generative models”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 127 records · Page 7Linked to original sources
PubMed1995

Susceptibility of different cell layers of the anterior and posterior part of the piriform cortex to electrical stimulation and kindling: comparison with the basolateral amygdala and "area tempestas".

Several lines of evidence suggest that the piriform cortex functions as a generator in the development and propagation of forebrain (limbic type) seizures, particularly in the kindling model of epilepsy. It is, however, not clear where, within the rather large piriform cortex region, the generator resides, and how much tissue is involved. Highly sensitive loci to chemical or electrical stimulation have been described both in the deep anterior and posterior parts of the piriform cortex. Furthermore, data from piriform cortex slice preparations indicated that epileptiform potentials originate in deep structures, particularly the endopiriform nucleus that underlies the piriform cortex. In the present study, in rats, we implanted stimulation and recording electrodes in various rostrocaudal locations of the piriform cortex and endopiriform nucleus, including the "area tempestas", i.e. a structure in the anterior part of the piriform cortex previously proposed to be critically involved in the generation of convulsive seizures of limbic origin. Within the piriform cortex, electrodes were aimed at different cellular layers of this structure. For comparison, additional animals received electrodes in different parts of the basolateral amygdala. A total of 19 different locations was obtained in this way. The susceptibility of these locations to electrical stimulation was characterized by determining the threshold for induction of afterdischarges. The afterdischarge threshold was lowest in layer III of the posterior piriform cortex and some locations in the endopiriform nucleus, whereas amygdala and "area tempestas" displayed higher values. In several animals, particularly those with electrodes in layer III of the posterior piriform cortex, spontaneous spiking was seen in prestimulation recordings, whereas this was never observed in recordings from the amygdala. Subsequent kindling by repeated stimulation of the various locations demonstrated marked differences in afterdischarge threshold reduction and kindling rate. The most marked decreases in afterdischarge threshold were seen in locations within layer III of the piriform cortex, whereas several other locations, including the "area tempestas", exhibited only moderate decreases or no decrease at all. In contrast to previous observations with only few locations in the piriform cortex region, the posterior piriform cortex was not in general slower to kindle than the anterior piriform cortex, although some locations in the posterior piriform cortex exhibited significantly lower kindling rates than the amygdala. The highest kindling rate was seen in the dorsal endopiriform nucleus.(ABSTRACT TRUNCATED AT 400 WORDS)

Amygdala↗
PubMed2001

Prediction of electroencephalographic spectra from neurophysiology.

A recent neurophysical model of propagation of electrical waves in the cortex is extended to include a physiologically motivated subcortical feedback loop via the thalamus. The electroencephalographic spectrum when the system is driven by white noise is then calculated analytically in terms of physiological parameters, including the effects of filtering of signals by the cerebrospinal fluid, skull, and scalp. The spectral power at low frequencies is found to vary as f(-1) when awake and f(-3) when asleep, with a breakpoint to a steeper power-law tail at frequencies above about 20 Hz in both cases; the f(-1) range concurs with recent magnetoencephalographic observations of such a regime. Parameter sensitivities are explored, enabling a model with fewer free parameters to be proposed, and showing that spectra predicted for physiologically reasonable parameter values strongly resemble those observed in the laboratory. Alpha and beta peaks seen near 10 Hz and twice that frequency, respectively, in the relaxed wakeful state are generated via subcortical feedback in this model, thereby leading to predictions of their frequencies in terms of physiological parameters, and of correlations in their occurrence. Subcortical feedback is also predicted to be responsible for production of anticorrelated peaks in deep sleep states that correspond to the occurrence of theta rhythm at around half the alpha frequency and sleep spindles at 3/2 times the alpha frequency. An additional positively correlated waking peak near three times the alpha frequency is also predicted and tentatively observed, as are two new types of sleep spindle near 5/2 and 7/2 times the alpha frequency, and anticorrelated with alpha. These results provide a theoretical basis for the conventional division of EEG spectra into frequency bands, but imply that the exact bounds of these bands depend on the individual. Three types of potential instability are found: one at zero frequency, another in the theta band at around half the alpha frequency, and a third at the alpha frequency itself.

Adult↗
PubMed2003

Using models to enhance the intellectual content of learning in developmental biology.

Models have been particularly useful in developmental biology over the last 30 years. At first, underlying control mechanisms were poorly understood, but over time a wealth of detailed information became available to provide an increasingly detailed knowledge of underlying mechanisms, at levels from genes through cells to organs, organisms and populations. Models are also of great value in teaching developmental biology, as they allow students to explore phenomena hard to perceive directly because of their scale, accessibility, expense or other considerations. A model may allow students to "experiment" in ways which would be impractical in real life, as well as give them a deep understanding of competing hypotheses of development. Lastly, students can be challenged to produce models of their own, whereas only rarely are they able to carry out original experiments. I discuss two main kinds of models and their uses in generating, testing and expounding hypotheses and point out dangers in the use of models in education. Models may draw upon and reflect the consensus paradigm in the field: a researcher may be able to appreciate that models are interim conditional statements of probability and use them to generate new knowledge. A student may be less able to do so and may fail to appreciate where new knowledge will come from. And unlike physics, biology is stochastic and contingent and can never be entirely deduced from first principles, implying that models can never be as perfect in any biological field as they can be in some other fields.

Developmental Biology↗
PubMed1997

Increased excitability and inward rectification in layer V cortical pyramidal neurons in the epileptic mutant mouse Stargazer.

The excitability of layer V cortical pyramidal neurons was studied in vitro in the single-locus mutant mouse stargazer (stg), a genetic model of spike wave epilepsy. Field recordings in neocortical slices from mutant mice bathed in artificial cerebrospinal fluid revealed spontaneous synchronous network discharges that were never present in wild-type slices. Intracellular and whole cell recordings from stg/stg neurons in deep layers showed spontaneous giant depolarizing excitatory post-synaptic potentials generating bursts of action potentials, and a 78% reduction in the afterburst hyperpolarization. Whole cell recordings revealed gene-linked differences in active membrane properties in two types of regular spiking neurons. Single action potential rise and decay times were reduced, and the rheobase current was decreased by 68% in mutant cells. Plots of spike frequency-current relationships revealed that the gain of this relation was augmented by 29% in the mutant. Comparisons of visually identified pyramidal neurons firing properties in both genotypes revealed no difference in single action potential afterhyperpolarization. Voltage-clamp recordings showed an approximately threefold amplitude increase in a cesium-sensitive inward rectifier. No cell density or soma size differences were observed in the layer V pyramidal neuron population between the two genotypes. These results demonstrate an autonomous increase in cortical network excitability in a genetic epilepsy model. This defect could lower the threshold for aberrant thalamocortical spike wave oscillations in vivo, and may contribute to the mechanism of one form of inherited absence epilepsy.

Action Potentials↗
PubMed2005

The physics of guiding catheters for the left coronary artery in transfemoral and transradial interventions.

BACKGROUND: The backup force of a guiding catheter is important for successful percutaneous coronary intervention (PCI), however, no theory has been proposed thus far regarding the factors involved in its generation. METHODS AND RESULTS: The backup force of guiding catheters was measured in an arterial tree model. In vitro modeling showed that larger-sized guiding catheters had greater backup force (8 Fr > 7 Fr > 6 Fr). Comparing the backup force between transfemoral (TFI) and transradial interventions (TRI), it was found to be 60% greater in TFI with a Judkins L (JL) catheter, and 8% greater in TFI with a backup (EBU/XB) type catheter. However, the Ikari L (IL) catheter generated a similar backup force between TRI and TFI. In TRI, the Ikari guiding catheter showed the greatest backup force, especially in the power position (power position of IL4 > IL4 > backup type 3.5 > deep engagement of JL4 > JL3.5 > JL4). These findings were associated with the angle of the catheter on the reverse side of the aorta. We then constructed several catheters with varying contact lengths. In vitro modeling showed that a longer contact area increased the backup force. CONCLUSIONS: The present model showed that three factors were associated with backup force: (1) catheter size; (2) angle on the reverse side of the aorta; and (3) contact area. The Ikari guiding catheter comprises all of the preferable factors in TRI.

Angioplasty, Balloon, Coronary↗
PubMed2006

Building new motor responses: eyelid conditioning revisited.

Neural processes underlying memory and learning should be studied under the best possible physiological conditions - namely, in alert behaving animals. The classical conditioning of the nictitating membrane and eyelid response is a widely used experimental model for studying the neural bases of motor learning in mammals. Nevertheless, information is still needed on the functional aspects, taking place simultaneously in different cerebral structures, that underlie acquisition, extinction and recall of new motor and cognitive abilities. Here, we review recent data on the neural activity generated in selected brain sites (facial motor nuclei, deep cerebellar nuclei and the hippocampus) in simultaneity with the process of learning. The use of modern technologies for the proper recording of eyelid movements, for the identification of the recorded units, and for the activation of selective synaptic processes during the learning situation enables a precise redefinition of the role played by these neural structures in such associative learning. This review is part of the TINS special issue on The Neural Substrates of Cognition.

Animals↗
PubMed2004

Electric field and stimulating influence generated by deep brain stimulation of the subthalamic nucleus.

OBJECTIVE: The goal of this project was to develop a quantitative understanding of the volume of axonal tissue directly activated by deep brain stimulation (DBS) of the subthalamic nucleus (STN). METHODS: The 3-dimensionally inhomogeneous and anisotropic tissue medium surrounding DBS electrodes complicates our understanding of the electric field and tissue response generated by the stimulation. We developed finite element computer models to address the effects of DBS in a homogeneous isotropic medium, and a medium with tissue conductivity properties derived from human diffusion tensor magnetic resonance data. The second difference of the potential distribution generated in the tissue medium was used as a predictor of the volume of tissue supra-threshold for axonal activation. RESULTS: The model predicts that clinically effective stimulation parameters (-3 V; 0.1 ms; 150 Hz) result in activation of large diameter (5.7 microm) myelinated axons over a volume that spreads outside the borders of the STN. The shape of the activation volume was dependent on the strong dorsal-ventral anisotropy of the internal capsule, and the moderate anterior-posterior anisotropy of the region around zona incerta. CONCLUSIONS: Small deviations ( approximately 1 mm) in the electrode position within STN can substantially alter the shape of the activation volume as well as its spread to neighboring structures. SIGNIFICANCE: STN DBS represents an effective treatment for medically refractory movement disorders such as Parkinson's disease. However, stimulation induced side effects such as tetanic muscle contraction, speech disturbance and ocular deviation are not uncommon. Quantitative characterization of the spread of stimulation will aid in the development of techniques to maximize the efficacy of DBS.

Artifacts↗
PubMed2005

In vivo determination of normal and anterior cruciate ligament-deficient knee kinematics.

The objective of the current study was to use fluoroscopy to accurately determine the three-dimensional (3D), in vivo, weight-bearing kinematics of 10 normal and five anterior cruciate ligament deficient (ACLD) knees. Patient-specific bone models were derived from computed tomography (CT) data. 3D computer bone models of each subject's femur, tibia, and fibula were recreated from the CT 3D bone density data. Using a model-based 3D-to-2D imaging technique registered CT images were precisely fit onto fluoroscopic images, the full six degrees of freedom motion of the bones was measured from the images. The computer-generated 3D models of each subject's femur and tibia were precisely registered to the 2D digital fluoroscopic images using an optimization algorithm that automatically adjusts the pose of the model at various flexion/extension angles. Each subject performed a weight-bearing deep knee bend while under dynamic fluoroscopic surveillance. All 10 normal knees experienced posterior femoral translation of the lateral condyle and minimal change in position of the medial condyle with progressive knee flexion. The average amount of posterior femoral translation of the lateral condyle was 21.07 mm, whereas the average medial condyle translation was 1.94 mm, in the posterior direction. In contrast, all five ACLD knees experienced considerable change in the position of the medial condyle. The average amount of posterior femoral translation of the lateral condyle was 17.00 mm, while the medial condyle translation was 4.65 mm, in the posterior direction. In addition, the helical axis of motion was determined between maximum flexion and extension. A considerable difference was found between the center of rotation locations of the normal and ACLD subjects, with ACLD subjects exhibiting substantially higher variance in kinematic patterns.

Adult↗
PubMed2004

Pattern formation in a metastable, gradient-driven sandpile.

With a toppling rule which generates metastable sites, we explore the properties of a gradient-driven sandpile that is minimally perturbed at one boundary. In two dimensions we find that the transport of grains takes place along deep valleys, generating a set of patterns as the system approaches the stationary state. We use two versions of the toppling rule to analyze the time behavior and the geometric properties of clusters of valleys, also discussing the relation between this model and the general properties of models displaying self-organized criticality.

Journal Article↗
PubMed2003

Polymerization contraction stresses of resin-based composite restorations within beveled cavity preparations of Class I restorations.

PURPOSE: To evaluate internal stresses resulting from polymerization of resin-based composite materials in cavity preparations beveled at the internal line angles or the cavosurface margin, compared with those in conventional butt-joint box-shaped cavity preparations. MATERIALS AND METHODS: Contraction stress generated in resin-based composite restorations placed in an experimental cavity preparation model was determined using micro-photoelastic analysis. Three types of cavity preparations simulating Class I restorations with bevels at internal line angles (1 mm and 2 mm deep), and at the cavosurface margin (2 mm deep) were prepared in a posterior composite model in order to obtain no marginal defect with the restorative composite. The transparent restorative composite was bulk-filled into the preparations and light-cured for 80 seconds. Specimens 2 mm thick for micro-photoelastic analysis were cut perpendicular to the long axis of the preparation. Fringe patterns for directions and magnitudes of stresses were obtained using transmitted and reflected polarized light. Then, the photoelastic stress analysis was performed to examine stresses in the preparations. The data were statistically compared with the previous data in conventional butt-joint box-shaped preparations by ANOVA and Scheffé's F test (P < 0.05). RESULTS: In 1 and 2 mm deep internal beveled cavity preparations, the maximum principal stress lines were parallel to the wall at the straight internal bevels. In the preparation with a round bevel at the cavosurface margin, the stress distribution in the bulk of the preparation was similar to that in the butt-joint preparation. Maximum stress values in 1 and 2 mm deep internal beveled preparations were 12.2 +/- 0.2 MPa and 18.8 +/- 2.8 MPa, respectively. For 2 mm-deep preparation, the maximum stress values in the internal beveled preparations were significantly lower than those values in butt-joint box-shaped cavity preparations. Maximum stress values in the preparations with the bevel at the cavosurface margin were 22.7 +/- 1.0 MPa, and were not significantly different from those values observed in the butt-joint preparations.

Analysis of Variance↗
PubMed2004

Model simulations of the Arctic atmospheric boundary-layer from the SHEBA year.

We present Arctic atmospheric boundary-layer modeling with a regional model COAMPS, for the Surface Heat Budget of the Arctic Ocean (SHEBA) experiment. Model results are compared to soundings, near-surface measurements and forecasts from the ECMWF model. The near-surface temperature is often too high in winter, except in shorter periods when the boundary layer was cloud-capped and well-mixed due to cloud-top cooling. Temperatures are slightly too high also during the summer melt season. Effects are too high boundary-layer moisture and formation of too dense stratocumulus, generating a too deep well-mixed boundary layer with a cold bias at the simulated boundary-layer top. Errors in temperature and therefore moisture are responsible for large errors in heat flux, in particular in solar radiation, by forming these clouds. We conclude that the main problems lie in the surface energy balance and the treatment of the heat conduction through the ice and snow and in how low-level clouds are treated.

Arctic Regions↗
PubMed2023

Active learning of enhancer and silencer regulatory grammar in photoreceptors.

Cis-regulatory elements (CREs) direct gene expression in health and disease, and models that can accurately predict their activities from DNA sequences are crucial for biomedicine. Deep learning represents one emerging strategy to model the regulatory grammar that relates CRE sequence to function. However, these models require training data on a scale that exceeds the number of CREs in the genome. We address this problem using active machine learning to iteratively train models on multiple rounds of synthetic DNA sequences assayed in live mammalian retinas. During each round of training the model actively selects sequence perturbations to assay, thereby efficiently generating informative training data. We iteratively trained a model that predicts the activities of sequences containing binding motifs for the photoreceptor transcription factor Cone-rod homeobox (CRX) using an order of magnitude less training data than current approaches. The model's internal confidence estimates of its predictions are reliable guides for designing sequences with high activity. The model correctly identified critical sequence differences between active and inactive sequences with nearly identical transcription factor binding sites, and revealed order and spacing preferences for combinations of motifs. Our results establish active learning as an effective method to train accurate deep learning models of cis-regulatory function after exhausting naturally occurring training examples in the genome.

Journal Article↗
PubMed2001

Organic carbon composition of marine sediments: effect of oxygen exposure on oil generation potential.

Anaerobic sedimentary conditions have traditionally been linked to the generation of the source rocks for petroleum formation. However, the influence of sedimentary redox conditions on the composition of freshly deposited organic matter (OM) is not clear. We assessed the effect of in situ exposure time to oxic conditions on the composition of OM accumulating in different coastal and deep-sea sediments using solid-state 13C nuclear magnetic resonance (NMR). 13C NMR spectra were resolved into mixtures of model components to distinguish between alkyl carbon present in protein and nonprotein structures. There is an inverse relation between the length of exposure to oxic conditions and the relative abundance of nonprotein alkyl (alkylNP) carbon, whose concentration is two orders of magnitude higher in coastal sediments with short exposure times than in deep-sea sediments with long exposure times. All alkylNP-rich samples contain a physically separate polymethylene component similar in composition to algaenans and kerogens in type I oil shales. The duration of exposure to oxic conditions appears to directly influence the quality and oil generation potential of OM in marine shales.

Journal Article↗
PubMedDate not supplied

Precise quantification of pulsatility is a necessity for direct comparisons of six different pediatric heart-lung machines in a neonatal CPB model.

Generation of pulsatile flow depends on an energy gradient. Surplus hemodynamic energy (SHE) is the extra hemodynamic energy generated by a pulsatile device when the adequate pulsatility is achieved. The objective of this study was to precisely quantify and compare pressure-flow waveforms in terms of surplus hemodynamic energy levels of six different pediatric heart-lung machines in a neonatal piglet model during cardiopulmonary bypass (CPB) procedures with deep hypothermic circulatory arrest (DHCA). Thirty-nine piglets (average weight, 3 kg) were subjected to CPB with a hydraulically driven physiologic pulsatile pump (PPP; n=7), Jostra-HL 20 pulsatile roller pump (Jostra-PR; n=6), Stockert Sill pulsatile roller pump (SIII-PR; n=6), Stockert Sill mast-mounted pulsatile roller pump with a miniature roller head (Mast-PR; n=7), Stockert Sill mast-mounted nonpulsatile roller pump (Mast-NP; n=7), or Stockert CAPS nonpulsatile roller pump (CAPS-NP, n=7). Once CPB was begun, each animal underwent 20 minutes of hypothermia, 60 minutes of DHCA, 10 minutes of cold reperfusion, and 40 minutes of rewarming. The pump flow rate was maintained at 150 ml x kg(-1) x min(-1) and the mean arterial pressure (MAP) at 45 mm Hg. In the pulsatile experiments, the pump rate was kept at 150 bpm and the stroke volume at 1 ml/kg. The SHE (ergs/cm3) = 1,332 ([(integral fpdt) / (integral fdt)] - MAP) was calculated at each experimental stage. During normothermic CPB (15 minutes on pump), the physiologic pulsatile pump generated the highest surplus hemodynamic energy (8563 +/- 1918 ergs/cm3, p < 0.001) compared with all other pumps. The Jostra HL-20 and Stockert Sill pulsatile roller pumps also produced adequate surplus hemodynamic energy. Nonpulsatile roller pumps and the Stockert Sill mast-mounted pulsatile roller pump did not generate any extra hemodynamic energy. During hypothermic CPB and after DHCA and rewarming, the results were extremely similar to those seen during normothermic CPB. The surplus hemodynamic energy formula is a novel method to precisely quantify different levels of pulsatility and nonpulsatility for direct and meaningful comparisons. The PPP produced the greatest surplus hemodynamic energy. Most of the pediatric pulsatile pumps (except Mast-PR) generated significant surplus hemodynamic energy. None of the nonpulsatile roller pumps generated adequate surplus hemodynamic energy.

Animals↗
PubMed2021

Deep generative neural network for accurate drug response imputation.

Drug response differs substantially in cancer patients due to inter- and intra-tumor heterogeneity. Particularly, transcriptome context, especially tumor microenvironment, has been shown playing a significant role in shaping the actual treatment outcome. In this study, we develop a deep variational autoencoder (VAE) model to compress thousands of genes into latent vectors in a low-dimensional space. We then demonstrate that these encoded vectors could accurately impute drug response, outperform standard signature-gene based approaches, and appropriately control the overfitting problem. We apply rigorous quality assessment and validation, including assessing the impact of cell line lineage, cross-validation, cross-panel evaluation, and application in independent clinical data sets, to warrant the accuracy of the imputed drug response in both cell lines and cancer samples. Specifically, the expression-regulated component (EReX) of the observed drug response achieves high correlation across panels. Using the well-trained models, we impute drug response of The Cancer Genome Atlas data and investigate the features and signatures associated with the imputed drug response, including cell line origins, somatic mutations and tumor mutation burdens, tumor microenvironment, and confounding factors. In summary, our deep learning method and the results are useful for the study of signatures and markers of drug response.

Antineoplastic Agents↗
PubMed2005

Crystal structure of the human urokinase plasminogen activator receptor bound to an antagonist peptide.

We report the crystal structure of a soluble form of human urokinase-type plasminogen activator receptor (uPAR/CD87), which is expressed at the invasive areas of the tumor-stromal microenvironment in many human cancers. The structure was solved at 2.7 A in association with a competitive peptide inhibitor of the urokinase-type plasminogen activator (uPA)-uPAR interaction. uPAR is composed of three consecutive three-finger domains organized in an almost circular manner, which generates both a deep internal cavity where the peptide binds in a helical conformation, and a large external surface. This knowledge combined with the discovery of a convergent binding motif shared by the antagonist peptide and uPA allowed us to build a model of the human uPA-uPAR complex. This model reveals that the receptor-binding module of uPA engages the uPAR central cavity, thus leaving the external receptor surface accessible for other protein interactions (vitronectin and integrins). By this unique structural assembly, uPAR can orchestrate the fine interplay with the partners that are required to guide uPA-focalized proteolysis on the cell surface and control cell adhesion and migration.

Crystallization↗
PubMed2001

Classical and quantum periodically driven scattering in one dimension.

Irregular scattering at harmonically driven one-dimensional potential wells is studied both on the classical and the quantum level. We show that an ac-driven single square well, and a smooth well with oscillating bottom, are sufficient to generate chaotic scattering. For a square well with oscillating bottom, we introduce the concept of pseudointegrable scattering. The quantum dynamics of these models is treated using Floquet scattering theory, which is exact for arbitrary amplitude and frequency of the driving. In the deep quantum regime, scattering is dominated by multiphoton exchanges with the driving field, leading to complex resonance structures in transmission and reflection. For strong and fast driving, the ac-driven square well develops an effective double-well potential that introduces coherent tunneling in the scattering. We identify signatures of classical chaotic scattering in a phase-space representation of the quantum dynamics.

Journal Article↗
PubMed2025

Smarter stomata: emergent technologies unlocking yield potential in a changing climate.

Stomata, the gatekeepers of leaf gas exchange, regulate carbon dioxide uptake and water loss, functions increasingly critical as crops face more frequent, intense heat and drought. Under dry conditions, stomatal conductance (g s) typically decreases, limiting carbon assimilation and yield. Heat stress, in contrast, elicits variable g S responses: sometimes increasing to facilitate transpirational cooling, while at other times decreasing, especially when combined with drought. Heat and drought also induce complex, context-dependent shifts in stomatal anatomy. Smaller, denser stomata improve drought resilience in some cases, while reduced density confers greater tolerance in others. The optimal stomatal ideotype remains unknown, and different or even opposing traits may confer resilience dependent on the environmental scenario. Substantial genotypic variation in g s and stomatal anatomy, high heritability and co-localized quantitative trait loci for stomatal traits and yield highlight their untapped potential as breeding targets for climate-resilient crops. However, stomatal traits remain largely absent from breeding pipelines due to challenges of phenotyping at scale. This is changing rapidly. Advances in deep learning, porometry, digital microscopy, and remote sensing now enable high-throughput measurement of stomatal physiology and anatomy. Next-generation breeding technologies including clustered regularly interspaced short palindromic repeats (CRISPR), multi-omics approaches, and artificial intelligence-driven ideotype selection models could revolutionize breeding, allowing precise engineering of stomatal traits for resilience to environmental stress. The time has come to move beyond characterizing stomatal traits and start actively incorporating them into breeding strategies. By leveraging these technologies, stomatal traits can become high value targets, unlocking their potential to enhance crop performance in a hotter, drier future.

abiotic stress↗