Volume 2 (2026) 

The research progress on the application of artificial intelligence in the identification of plant calcium-related genes and the regulation of signaling pathways

Citation

Jia F, Wang M. The research progress on the application of artificial intelligence in the identification of plant calcium-related genes and the regulation of signaling pathways. Engineering Innovation and Practice, 2026, 2, eip2v0118a. 

Abstract

       Calcium ions (Ca²⁺), as crucial second messengers in plant cells, play a central role in growth, development, metabolic regulation, and stress responses. Systematic analysis of calcium-related genes and signaling pathways is fundamental to understanding calcium regulatory mechanisms and enhancing crop stress resistance. In recent years, with the rapid accumulation of multi-omics data—including genomics, transcriptomics, proteomics, and metabolomics—the application of artificial intelligence (AI) in plant molecular network analysis has advanced significantly, providing new insights for identifying complex signaling pathways and optimizing metabolic networks. This review summarizes recent progress in AI-based identification of calcium-related genes and modeling of signaling pathways, with a particular focus on the application of algorithms such as graph neural networks, convolutional neural networks, and random forests in constructing calcium signaling networks and mining key gene modules. Moreover, it discusses the potential of AI in reconstructing calcium-regulated metabolic dynamics, elucidating stress-responsive signaling patterns, and improving calcium use efficiency and metabolic homeostasis in crops. Finally, the challenges of data integration, algorithm interpretability, and cross-scale modeling are analyzed, and the future directions of AI-driven calcium signaling research in precision agriculture and intelligent breeding are outlined. This review aims to provide a systematic theoretical reference and technical framework for plant calcium signaling studies and crop metabolic optimization.

Keywords

     artificial intelligence in plant biology, calcium signaling pathway, calcium-related genes, graph neural network, crop stress resistance, multi-omics integration, metabolic network modeling, precision agriculture

1. Introduction

Calcium ions (Ca²⁺), as essential inorganic elements in plants, perform multiple physiological functions during growth, development, and environmental responses. Beyond serving as vital components of the cell wall and membrane structure, Ca²⁺ is widely recognized as a pivotal second messenger occupying a central role in cellular signaling networks, regulating diverse biological processes such as cell division, elongation, reproductive development, and organ formation [1,2]. In stress responses, calcium mediates plant adaptation to drought, salinity, low temperature, and pathogen infection by modulating ion channel activity, enzymatic reaction rates, and gene expression patterns [3,4]. Therefore, elucidating the dynamic regulatory mechanisms of calcium signaling in plants is of fundamental theoretical and practical significance for understanding environmental adaptability and improving crop stress resistance. 

In recent years, substantial progress has been made in the study of plant calcium signaling pathways. Research indicates that transient fluctuations in intracellular Ca²⁺ concentrations generate unique “calcium signatures,” which are decoded by calcium sensors such as calmodulins (CaMs), calcium-dependent protein kinases (CDPKs), and calmodulin-like proteins (CMLs), subsequently transmitting signals to downstream effectors and triggering transcriptional and metabolic responses [5-7]. However, the molecular mechanisms of calcium signaling are complex and multifaceted, characterized by spatiotemporal regulation and extensive pathway crosstalk, with significant differences observed among species, tissues, and developmental stages[8-10]. Although partial calcium signaling networks have been constructed, the overall regulatory framework and key genetic nodes remain incompletely understood, particularly concerning their functional roles in crop stress adaptation and metabolic regulation. 

With the rapid development of high-throughput omics and systems biology, massive datasets from genomics, transcriptomics, and metabolomics have provided unprecedented resources for calcium signaling research. Nevertheless, the complexity and heterogeneity of these datasets present substantial challenges for traditional analytical methods. The recent introduction of artificial intelligence (AI) technologies has opened new avenues for systematic modeling and functional interpretation of plant signaling networks [11]. AI leverages machine learning, deep learning, and graph neural network (GNN) algorithms to efficiently mine potential gene regulatory patterns from multidimensional omics data, identify key regulatory modules, and reconstruct signaling pathways [12,13]. For instance, random forest and convolutional neural network (CNN) models have been applied to predict gene expression patterns and functional associations under stress conditions [14], while GNN-based models can identify critical nodes and boundary relationships in complex molecular interaction graphs of calcium signaling [15]. Furthermore, AI-assisted approaches for metabolic network optimization and feature extraction are increasingly employed in studies of calcium absorption, transport, and redistribution in crops, offering novel strategies for precision nutrient management and intelligent breeding [16,17]. 

In this context, the present review aims to systematically summarize the latest research progress on calcium-related gene identification and signaling pathway regulation from the perspective of artificial intelligence. It explores the potential of AI algorithms in multi-omics data integration, signaling network modeling, and metabolic optimization. By highlighting methodological innovations and application trends of AI in plant calcium signaling research, this review seeks to provide theoretical insights and technological support for crop stress resilience improvement, nutrient efficiency enhancement, and the advancement of intelligent agriculture. 

2. Identification and analysis of calcium-related genes 

2.1. Integration of genomic and transcriptomic data 

The integration of genomic and transcriptomic data provides a fundamental basis for the systematic identification and functional characterization of calcium-related genes. Genomics offers a comprehensive genetic blueprint of plants, encompassing gene sequences, locations, and potential functions. Through high-quality whole-genome sequencing and annotation, researchers have identified gene families closely associated with calcium signaling, such as CaMs, CDPKs, CMLs, and calcium sensor proteins [18,19]. These genes play central roles in calcium perception, transduction, and cellular responses. Analyses of their genomic structure, copy number variation, and cis-regulatory elements provide insights into their evolutionary patterns and functional diversity. Furthermore, genomic information establishes a theoretical foundation for functional prediction and phylogenetic analysis of gene families, serving as a critical reference for subsequent transcriptomic studies. 

Guided by genomic information, transcriptomic analysis reveals the dynamic expression patterns of calcium-related genes across tissues, developmental stages, and environmental stresses. Using high-throughput RNA sequencing (RNA-Seq), researchers can systematically assess gene expression changes under specific stimuli such as calcium treatments, drought, salinity, and low temperature [20]. Transcriptomic data not only provide quantitative expression profiles but also uncover gene co-expression patterns and regulatory networks. For instance, in model plants such as Arabidopsis thaliana and Oryza sativa, CDPK and CaM genes associated with Ca²⁺ signaling pathways are significantly upregulated under salt stress, suggesting their crucial roles in ion homeostasis regulation and stress adaptation [21]. Therefore, integrating transcriptomic data with genomic information enables the identification of key regulatory genes involved in calcium signaling and provides candidate targets for functional validation. 

The integration of genomic and transcriptomic analyses typically relies on various statistical and computational biology approaches, including correlation analysis, principal component analysis (PCA), and weighted gene co-expression network analysis (WGCNA) [22]. These methods elucidate the potential relationships among different data layers and identify core regulatory factors influencing calcium signaling under specific physiological or environmental conditions. In recent years, the incorporation of AI and machine learning approaches has further enhanced analytical precision. Algorithms such as random forests, GNNs, and Bayesian networks have been employed to construct high-dimensional gene co-expression networks, enabling the precise identification of key gene modules and cooperative regulatory mechanisms in calcium signaling pathways [23,24]. Such multi-omics integration strategies not only deepen the systematic understanding of calcium-related gene functions and regulatory networks but also provide new scientific foundations for elucidating plant adaptive mechanisms, optimizing crop breeding, and improving calcium nutrient management. 

2.2. Screening of calcium signaling-related genes

The screening of calcium signaling-related genes represents a critical step in elucidating the molecular mechanisms underlying plant calcium regulatory networks. This research relies on the deep integration of bioinformatics tools and high-throughput omics technologies. Through systematic analysis of plant genome sequences, researchers can identify conserved gene regions and sequence motifs closely associated with calcium signaling pathways. These genes typically include members of families encoding CaMs, CDPKs, and CMLs [25-27]. They play central roles in the perception and transduction of calcium signals and are key factors in maintaining cellular homeostasis and mediating environmental responses. To enhance the systematicity and accuracy of gene screening, homology-based sequence alignment and hidden Markov model (HMM) algorithms are often employed to identify conserved calcium signaling pathway genes across different species, thereby accelerating the prediction of unknown gene functions and the identification of candidate genes.

In addition to genomic sequence alignment, differential expression analysis using transcriptomic data is another essential approach for screening calcium signaling-related genes. High-throughput RNA sequencing (RNA-seq) enables systematic profiling of dynamic gene expression patterns under diverse physiological and environmental conditions, such as calcium treatment, salinity, drought, low temperature, and pathogen infection [28]. Differentially expressed gene (DEG) analysis reveals genes that are significantly upregulated or downregulated in response to calcium signaling, providing clues to their roles in signal perception, transduction, and response processes [29]. For instance, recent studies have shown that multiple CDPK genes in Arabidopsis thaliana and maize exhibit markedly elevated expression under salt stress and calcium stimuli, suggesting their crucial roles in ion homeostasis and stress regulation. By integrating diverse transcriptomic datasets, researchers can not only construct a preliminary framework of calcium signaling regulatory networks but also identify target genes for downstream functional validation.

In recent years, the application of machine learning and AI algorithms has provided novel strategies for screening calcium signaling-related genes. By integrating genomic, transcriptomic, and proteomic data into machine learning models, multidimensional regulatory networks of calcium signaling can be constructed to uncover potential key genes and regulatory modules [30-32]. For example, random forest and support vector machine (SVM) algorithms have been employed to extract gene features strongly associated with calcium signaling responses from high-dimensional omics data, while deep learning approaches such as GNNs can model complex gene–gene interactions and reveal hierarchical relationships and topological structures within signaling pathways. These AI-driven, data-intensive analytical frameworks greatly enhance the efficiency and precision of calcium signaling gene screening, providing a robust computational foundation for in-depth exploration of plant calcium signaling networks and functional genomics.

2.3. Identification of key gene expression patterns 

The identification of key gene expression patterns is a crucial step in elucidating the core regulatory mechanisms of calcium signaling pathways. During calcium signal transduction, gene expression patterns are often closely associated with specific physiological processes or environmental stimuli [33,34]. Therefore, by comparing gene expression differences under various conditions, researchers can infer the functional characteristics of genes involved in plant growth, development, and stress responses. Using high-throughput omics techniques such as RNA sequencing (RNA-seq), genome-wide expression profiles can be obtained under experimental conditions including exogenous Ca²⁺ stimulation, drought, salinity, and low temperature. DEG analysis enables the identification of genes that are significantly upregulated or downregulated during calcium signaling responses. These genes typically represent key regulatory nodes within the calcium signaling cascade and may participate in calcium perception, signal transduction, and downstream response regulation. 

To further elucidate the regulatory mechanisms of calcium signaling-related genes, co-expression analysis has become an essential approach for revealing gene co-regulatory relationships. WGCNA allows researchers to identify gene modules exhibiting synchronized expression patterns during calcium signaling and infer their potential biological functions. Such co-expression networks facilitate the discovery of functionally related gene clusters and reveal hub genes and their interactions with essential regulators, such as CaMs and CDPKs [35,36]. For instance, in Arabidopsis thaliana and Oryza sativa, co-expression analyses have demonstrated that the synergistic interaction between CDPKs and the CBL–CIPK signaling complex plays a pivotal role in mediating stress responses and regulating ion homeostasis. This network-based analytical framework helps extract core regulatory modules from complex datasets, providing valuable candidate targets for functional validation. 

In addition, time-series transcriptomic analysis offers a novel perspective for uncovering the dynamic features of calcium signaling regulation. By collecting transcriptome data from plant tissues or cells at multiple time points following signal induction, researchers can map temporal trajectories of gene expression during calcium signal transduction, distinguishing between early perception genes and late response genes. Such dynamic expression analyses reveal hierarchical regulatory relationships at different stages of calcium signaling and clarify its spatiotemporal coordination in stress adaptation and plant development. In recent years, AI-assisted temporal modeling approaches—such as dynamic Bayesian networks and deep learning algorithms—have been introduced into plant signal transduction research, significantly enhancing the ability to identify complex transcriptional regulatory patterns [37-39]. These approaches are of great significance for understanding how calcium signaling enables rapid responses and long-term adaptation, thereby elucidating the molecular basis of plant environmental perception. 

3. Identification and network reconstruction of calcium signaling pathways 

3.1. Fundamental mechanisms of calcium signaling pathways 

Calcium signaling pathways play a central regulatory role in plant cells, acting as a pivotal transduction system for perceiving and responding to environmental stimuli. Ca²⁺, as crucial secondary messengers, participate extensively in plant growth, development, stress responses, and metabolic regulation [40]. When exposed to external factors such as changes in light intensity, temperature fluctuations, pathogen invasion, or drought and salt stress, the intracellular free Ca²⁺ concentration undergoes rapid transient fluctuations, generating characteristic calcium “spikes” or “waves” [41,42]. These Ca²⁺ signals are perceived and transmitted to downstream targets by calcium sensors and transducers, including CaMs, CDPKs, and calcium-binding proteins (CMLs). Through specific interactions with target proteins, these molecules mediate precise signaling cascades that regulate gene expression and metabolic processes. 

Calcium signaling is typically initiated by the activation of Ca²⁺ channels located on the plasma membrane or organellar membranes. Various types of calcium channels, such as mechanosensitive channels, cyclic nucleotide-gated channels, and voltage-dependent channels, are activated under external stimuli, allowing Ca²⁺ influx from the apoplast or organelles into the cytosol, thereby triggering a transient elevation in cytosolic Ca²⁺ concentration. Subsequently, calcium-sensing proteins rapidly detect changes in Ca²⁺ levels and translate them into phosphorylation cascades through interactions with downstream molecules. This process involves the coordinated action of signaling modules such as CaMs, CDPKs, and the calcineurin B-like protein–CIPK (CBL–CIPK) complex [43,44]. These signaling components interconnect to form multilayered and dynamic regulatory networks that integrate with hormone signaling and redox signaling pathways, thereby achieving fine-tuned control of plant developmental and adaptive responses. 

A distinctive feature of calcium signaling pathways is their “encoded” regulatory mechanism. Different frequencies, amplitudes, and durations of Ca²⁺ signals convey distinct physiological meanings, a phenomenon known as the “calcium signature” [45-47]. For instance, low-amplitude and short-period Ca²⁺ oscillations are often associated with growth regulation, whereas high-frequency and sustained signals are typically linked to stress defense responses. Plant cells decode these specific calcium signatures via “calcium decoder” proteins that recognize unique temporal and spatial signal characteristics, enabling rapid and precise physiological adaptation to environmental fluctuations. This spatiotemporal encoding–decoding mechanism provides a molecular foundation for plants to achieve high-fidelity regulation under complex and variable environmental conditions.

3.2. AI-based methods for gene network construction

The construction of gene regulatory networks is a critical step in elucidating the mechanisms underlying plant calcium (Ca²⁺) signaling. The introduction of AI technologies has markedly improved the efficiency, accuracy, and scalability of this process. Traditional approaches to gene network construction rely heavily on experimental data accumulation and manual inference, which are often time-consuming, costly, and limited in handling complex multi-omics data. With advances in high-throughput sequencing and big data analytics, AI—particularly machine learning (ML) and deep learning (DL) algorithms—has emerged as a powerful tool for multi-omics data analysis [48,49]. AI can automatically mine potential regulatory relationships from genomic, transcriptomic, metabolomic, and proteomic datasets, identify key genes and interaction patterns in Ca²⁺ signaling pathways, and provide an efficient and systematic framework for understanding calcium-regulated gene networks in plants.

Among various AI algorithms, GNNs have recently been widely applied to modeling complex biological networks. In GNN frameworks, genes are represented as nodes and their regulatory or interaction relationships as edges, forming multilayered gene interaction graphs [50]. Through iterative training, GNNs can automatically learn structural features from nonlinear gene–gene interactions and identify core regulatory genes and critical modules within Ca²⁺ signaling networks. Moreover, the strong generalization ability of GNNs enables the prediction of potential functions for unannotated genes and the discovery of novel signal transduction routes [51]. Compared with traditional statistical methods, GNNs more effectively capture the dynamic and hierarchical nature of calcium signaling networks, making them a powerful computational tool for deciphering complex gene regulatory architectures.

In addition to deep learning, classical machine learning models—such as random forest and SVM—have also demonstrated strong performance in identifying calcium-related genes and reconstructing regulatory networks. These models integrate multi-omics datasets (e.g., gene expression, protein–protein interactions, and metabolite profiles) and employ feature selection and cross-validation strategies to reduce data noise and pinpoint significant calcium-regulated genes [52,53]. Random forest models can rank gene importance to determine central regulatory nodes, while SVMs excel in handling high-dimensional and nonlinear data, accurately classifying calcium response patterns [54,55]. In recent years, AI-based models have also been applied to predict calcium nutrient utilization and stress resilience in crops, offering new computational tools and theoretical foundations for the systematic analysis of calcium signaling regulation and the development of molecular breeding strategies.

3.3. Calcium signaling-based mechanisms of plant responses to environmental stress

Calcium signaling plays a central regulatory role in plant adaptation to environmental stresses, particularly under drought, salinity, low temperature, and pathogen attack. As a pivotal second messenger, Ca²⁺ rapidly perceives external stimuli and triggers cellular defense responses. Upon exposure to stress signals, Ca²⁺ channels located on the plasma membrane and organellar membranes (such as the vacuole, endoplasmic reticulum, and plastids) open swiftly, resulting in a transient rise in cytosolic free Ca²⁺ concentration and the generation of characteristic “calcium waves” or “calcium spikes” [56,57]. This spatiotemporal dynamic Ca²⁺ signal acts as a primary sensing mechanism that converts external environmental cues into intracellular signaling events, subsequently activating cascades that regulate gene expression, post-translational modifications, and metabolic reprogramming, thereby enhancing plant adaptability to stress.

Under stress conditions, plants employ a variety of calcium-sensing and signaling molecules to detect and transduce calcium signals. CaM and CDPKs are key components of calcium signaling pathways that decode cytosolic Ca²⁺ fluctuations and relay signals to downstream effectors. CaM interacts with multiple transcription factors, such as members of the CAMTA family, to regulate the expression of stress-responsive genes [58], whereas CDPKs modulate the activity of osmotic regulators, antioxidant enzymes, and key enzymes in signaling molecule biosynthesis via phosphorylation [59]. Moreover, extensive cross-talk exists between calcium signaling and other signaling pathways, including reactive oxygen species (ROS), plant hormones (ABA: abscisic acid, JA: jasmonic acid, and mitogen-activated protein kinase (MAPK) cascades[60,61]. This networked signaling coordination enables plants to achieve integrated and systemic resistance responses under multifactorial stress conditions.

The precise transmission of calcium signals depends not only on the activation of calcium sensors but also on the spatiotemporal dynamics of calcium signaling patterns. The frequency, amplitude, and duration of calcium signals are critical parameters for cellular information “encoding,” as distinct signal features elicit specific physiological responses [62]. For example, transient Ca²⁺ spikes primarily mediate stomatal closure and ion homeostasis, whereas sustained or periodic calcium oscillations participate in transcriptional reprogramming and metabolic optimization, promoting long-term stress adaptation. This “calcium signature” encoding mechanism endows plants with high specificity and regulatory flexibility, allowing them to dynamically balance growth and defense under fluctuating environmental conditions.

4. AI-based gene network analysis

4.1. Application of AI algorithms in gene network construction

With the rapid advancement of multi-omics technologies, the volume of genomic, transcriptomic, and metabolomic data has grown exponentially. Traditional gene network construction methods, which rely on empirical modeling and experimental data accumulation, face limitations in efficiently analyzing such complex and multidimensional datasets. The introduction of AI, particularly machine learning (ML) and deep learning (DL) algorithms, has established a new paradigm for gene network reconstruction. AI algorithms can integrate omics data across different layers through nonlinear modeling and pattern recognition, thereby uncovering potential gene–gene associations and regulatory mechanisms. AI-driven network reconstruction not only accelerates the identification of key genes in calcium signaling pathways but also reveals hierarchical regulatory relationships among genes, deepening the systemic understanding of plant gene regulation under environmental stresses.

4.1.1. Application of graph neural networks

GNNs, as advanced AI-driven network learning models, are particularly well-suited for analyzing highly complex interactions among genes in biological systems. By representing genes as nodes and regulatory relationships as edges, GNNs construct high-dimensional gene regulatory graphs and iteratively learn latent features and dynamic patterns through multilayer neural training [63,64]. Compared with traditional correlation-based approaches, GNNs capture nonlinear dependencies and contextual interactions among genes, making them highly effective for elucidating hierarchical regulatory structures in calcium signaling pathways. In plant calcium signaling research, GNNs can identify key regulatory nodes, infer the directionality of gene interactions, and uncover the crosstalk between calcium signaling and other networks, such as hormone or ROS signaling, enabling system-level reconstruction of cellular regulation.

Furthermore, GNNs possess strong self-supervised learning capabilities, allowing them to extract latent functional modules from large-scale unlabeled omics datasets [65]. Given the dynamic and condition-dependent nature of plant calcium signaling networks—characterized by tissue specificity and temporal variation—GNNs demonstrate outstanding performance in integrating heterogeneous, high-dimensional data. Through GNN-based modeling, researchers can predict the functional roles of previously unannotated genes in calcium signaling, visualize the dynamic behavior of regulatory pathways, and achieve fine-scale modeling. These advances provide a robust computational foundation for future systems biology studies and AI-assisted crop improvement.

4.1.2. Application of random forest models

Random forest is an ensemble learning algorithm that has been widely employed in gene selection and network construction due to its high efficiency and robustness. By aggregating multiple decision trees through voting or averaging, random forests effectively handle nonlinear, high-dimensional, and noisy datasets. In studies of calcium signaling pathways, random forest models utilize multi-omics data to identify core genes strongly associated with calcium signaling and apply feature importance ranking to quantify each gene’s regulatory weight, thereby delineating the key nodal structure of calcium regulatory networks [66]. The random forest algorithm not only minimizes overfitting but also enhances the stability and accuracy of gene selection through cross-validation mechanisms.

The advantages of random forest in calcium signaling analysis also lie in its interpretability and compatibility with hybrid modeling frameworks. Researchers can integrate random forest with deep learning models—such as GNNs or CNNs—to achieve multi-algorithm optimization, improving both predictive power and biological interpretability [67]. For example, random forest can be employed to pre-screen potential calcium-regulatory genes, while GNNs subsequently perform network topology learning, forming a closed-loop framework of feature extraction and structural inference. This AI-integrated strategy significantly enhances gene identification accuracy and provides an effective computational tool for elucidating calcium signaling regulation in plant growth, development, and stress adaptation.

4.2. Identification and functional prediction of gene modules

The identification of gene modules is a key step in elucidating the structural and functional characteristics of gene networks. A gene module generally refers to a group of genes that exhibit co-expression or co-regulation patterns within specific biological processes or signaling pathways, often sharing similar functions or being controlled by the same regulatory mechanisms. In calcium signaling studies, identifying calcium-related gene modules helps reveal how different genes coordinate their activities within the signaling network and clarifies their specific roles in plant growth, development, and stress responses [68]. By integrating AI algorithms with co-expression analysis, researchers can automatically extract key modules from large-scale omics datasets and further analyze their roles in calcium signaling pathways, thereby uncovering the hierarchical structure of calcium signaling regulatory networks.

AI algorithms have demonstrated significant advantages in gene module identification. Clustering algorithms (such as K-means and hierarchical clustering) and unsupervised learning models can classify high-dimensional gene co-expression data into several functionally related modules. In particular, the integration of WGCNA and GNNs has greatly improved both the precision and interpretability of module identification [69,70]. WGCNA constructs weighted networks based on expression correlations between genes to identify groups that show coordinated changes in calcium signaling regulation, while GNNs capture nonlinear relationships and cross-module interaction patterns [71]. In plant calcium signaling research, these approaches have uncovered modular coordination among calcium sensor proteins (such as CaM and CMLs), CDPKs, and their downstream responsive genes, providing new insights into the dynamic regulation of calcium signaling networks. Moreover, AI models effectively integrate cross-species and cross-condition data, revealing the conservation and diversity of calcium signaling modules under varying ecological environments [72].

Following module identification, functional prediction becomes crucial for elucidating the biological significance of the identified modules. Deep learning and machine learning models (such as Random Forest and Support Vector Machines) can learn characteristic patterns of genes with known functions from large-scale training datasets and subsequently predict the potential roles of unknown genes within newly identified modules [73]. For instance, by incorporating gene expression profiles, transcription factor binding sites, and protein–protein interaction data, AI algorithms can predict the biological functions of calcium-related modules in processes such as cell wall remodeling, ion homeostasis, and stress response. This strategy significantly accelerates the screening and validation of functional genes and provides experimentally testable molecular targets. AI-driven functional prediction not only deepens the systematic understanding of calcium signaling regulatory mechanisms but also offers a theoretical basis for improving plant stress tolerance and advancing molecular breeding.

4.3. Key regulatory factors in calcium signaling pathways

In calcium signaling pathways, key regulatory factors refer to the genes and proteins that play central roles in the perception, transmission, amplification, and regulation of calcium signals [74]. Identifying these factors is essential for understanding how calcium signaling achieves precise intracellular transduction and how it participates in biological processes such as cell growth, differentiation, and stress responses. Typical regulatory factors include CaM, CDPKs, calcineurin B-like protein-interacting protein kinases (CIPKs), phosphatases, and transcription factors. These molecules sense dynamic fluctuations in intracellular calcium concentrations and translate external stimuli into specific transcriptional and metabolic responses, thereby ensuring accurate calcium signal regulation. Notably, these regulators often operate through complex feedback loops to achieve spatiotemporal specificity, maintaining a balance between signal intensity and duration.

CaM and CDPKs are among the most critical regulatory components in plant calcium signaling networks. CaM undergoes conformational changes upon binding with Ca²⁺, subsequently activating downstream target proteins such as CaM-dependent kinases and transcriptional regulators to amplify and transmit the signal. CDPKs, which possess dual calcium-sensing and effector functions, directly respond to calcium signals and modulate metabolic pathways, gene expression, and physiological processes through phosphorylation. Studies have shown that CDPKs play vital roles in multiple stress responses, including drought, salinity, and pathogen defense, by regulating stomatal movement, osmotic balance, and cell wall remodeling [75,76]. The interplay among these regulatory factors forms a highly dynamic and intricate signaling network that enables plants to respond rapidly and precisely to environmental fluctuations.

In recent years, the integration of AI technologies has greatly advanced the identification and functional characterization of calcium signaling regulators. By integrating multi-omics datasets (transcriptomics, proteomics, and metabolomics), machine learning and deep learning algorithms can efficiently screen regulatory factors closely associated with calcium signaling [77,78]. For instance, Random Forest and SVM models can identify genes showing significant differential expression under various stress conditions and extract potential key regulatory elements, while GNNs can construct interaction graphs to reveal core nodes within the calcium signaling network. These AI-driven approaches not only enhance the efficiency and accuracy of key factor identification but also enable the prediction of putative regulatory components that have not yet been experimentally validated, providing new insights and technological support for elucidating the complex regulatory mechanisms of calcium signaling networks.

5. Metabolic pathway reconstruction and optimization 

5.1. The central role of calcium in metabolic pathways 

Ca²⁺, as one of the most crucial secondary messengers in plant cells, plays a central role in the regulation of metabolic pathways. They not only mediate plant growth and development but also participate in diverse signal transduction and metabolic regulation processes [79]. Calcium signaling can rapidly respond to environmental stimuli, transmitting signals from the plasma membrane to the nucleus and metabolic systems, thereby triggering a cascade of enzymatic reactions and metabolic reprogramming. Dynamic fluctuations in intracellular Ca²⁺ concentrations directly influence the activities of metabolic enzymes and the flux of metabolic intermediates, exerting significant control over energy metabolism, substance transport, and stress adaptation. For instance, under salinity, drought, or low-temperature stresses, calcium signals perceive environmental cues and regulate metabolic rates, enabling plants to perform rapid metabolic reconfiguration to adapt to changing environments. 

Ca²⁺ primarily regulates the activities of key enzymes in metabolic pathways through CDPKs and their downstream effectors, achieving multilayered metabolic control [80]. These enzymes are widely involved in carbohydrate, lipid, and secondary metabolic pathways. For example, under drought stress, calcium signaling activates CDPK-mediated phosphorylation, regulating the synthesis of osmolytes such as proline and polyols to maintain cellular osmotic balance. Meanwhile, it enhances the activities of antioxidant enzymes, including superoxide dismutase (SOD), catalase (CAT), and ascorbate peroxidase (APX), to mitigate oxidative damage. Thus, calcium signaling not only modulates metabolic fluxes but also maintains metabolic homeostasis and physiological stability under adverse conditions. 

It is noteworthy that the regulatory function of calcium signaling in metabolism does not operate independently but is integrated into a complex signaling network involving hormones and ROS [81]. The crosstalk between calcium and phytohormone signaling, such as ABA and ethylene (ET), allows coordinated regulation of plant growth, development, and stress adaptation. Moreover, calcium signaling modulates the biosynthesis of numerous secondary metabolites—including polyphenols, flavonoids, and alkaloids—that play essential roles in plant defense, antioxidation, and environmental adaptation [82,83]. Through the synergistic interactions between calcium and other signaling pathways, plants achieve dynamic allocation and optimal utilization of metabolic resources across developmental stages and environmental conditions, thereby sustaining their growth, development, and physiological homeostasis. 

5.2. AI-based analysis of metabolomics data and pathway reconstruction

Metabolomics provides essential insights into the regulatory mechanisms through which calcium signaling participates in plant metabolism. By systematically analyzing the dynamic variations of metabolites, researchers can gain a deeper understanding of the functional roles of Ca²⁺ in metabolic regulation [84]. However, metabolomic datasets are typically high-dimensional, complex, and highly dynamic, posing significant challenges for traditional statistical and modeling approaches to extract underlying biological patterns. The integration of AI technologies—particularly machine learning and deep learning—has opened new avenues for intelligent analysis and metabolic pathway reconstruction [85]. Leveraging AI models enables efficient identification of Ca²⁺-associated metabolic pathways, the construction of complex metabolic networks, and the elucidation of metabolic flux dynamics and interactions under varying physiological conditions.

Applications of AI algorithms in metabolomics mainly involve pattern recognition, feature extraction, and multi-omics data integration. Deep learning models, such as CNNs and autoencoders, can learn latent structural features from metabolomic data, thereby identifying key metabolic modules and pathways regulated by calcium signaling [86]. For instance, AI-based clustering and dimensionality reduction can automatically reveal metabolic patterns under different stress conditions, uncovering how Ca²⁺ signaling regulates carbon metabolism, amino acid biosynthesis, and secondary metabolic pathways. Furthermore, AI-driven data fusion can integrate metabolomics with transcriptomics and proteomics, achieving cross-scale metabolic network reconstruction. This multidimensional integration enhances the systemic understanding of calcium-mediated metabolic networks and provides robust support for modeling complex physiological processes [87].

Through AI-driven metabolic pathway reconstruction, researchers can not only identify central nodes of calcium regulation in plant metabolism but also optimize metabolic network design to predict and validate calcium signaling mechanisms under specific physiological contexts [88]. For example, under drought or salt stress, AI algorithms can identify Ca²⁺-responsive pathways such as proline biosynthesis and antioxidant metabolism, predicting the dynamic influence of calcium signaling on metabolite accumulation. Moreover, AI-enabled metabolic optimization can guide precise calcium regulation strategies in crops by predicting the effects of varying Ca²⁺ supply levels on metabolic homeostasis and stress tolerance. Consequently, the deep integration of AI technologies is accelerating the systematic decoding and dynamic optimization of calcium-related metabolic networks, providing novel insights and technical foundations for plant metabolic engineering and stress-resilience improvement.

5.3. Regulation mechanisms of calcium signaling in the plant cell wall structure 

Ca²⁺ plays a central regulatory role in the formation, remodeling, and functional maintenance of plant cell walls. The plant cell wall not only maintains cellular morphology and provides mechanical support and a protective barrier, but also serves as an important regulatory platform for responding to environmental stresses. Calcium predominantly exists in the cell wall in the form of calcium–pectin complexes, forming “egg-box” structures through electrostatic cross-linking between pectin molecules, thereby enhancing the mechanical strength and resilience of the cell wall [89]. This calcium–pectin network not only strengthens the structural stability of the cell wall but also plays a critical role in cell expansion and stress responses. For instance, under salt stress or pathogen invasion, Ca²⁺ dynamically modulates the degree of pectin cross-linking, altering the porosity and extensibility of the cell wall and ultimately influencing the plant’s efficiency in stress adaptation. 

Calcium signaling regulates the dynamic reconstruction and repair of the cell wall by activating the expression and activity of wall-associated enzymes. CDPKs act as key mediators in Ca²⁺ signaling transduction, sensing cytosolic Ca²⁺ fluctuations and modulating the activity of cellulose synthase, pectin methylesterase, and polysaccharide cross-linking enzymes [90]. These regulatory effects facilitate the ordered synthesis and cross-linking of major wall components such as cellulose, hemicellulose, and pectin. Under stress conditions, Ca²⁺ signaling rapidly activates cell wall repair pathways by regulating the expression of degradative enzymes, including polygalacturonases and xylanases, to repair damaged wall regions. Furthermore, Ca²⁺ interacts with other signaling molecules, such as ROS and ABA, to form an intricate signaling network that coordinates wall homeostasis and defense mechanisms [91]. 

In addition, calcium signaling modulates cell expansion and growth direction by influencing the coupling between the cell wall and the plasma membrane. Cell elongation relies on reversible modulation of wall flexibility, and Ca²⁺ regulates the ionic microenvironment and hydration state within the wall, affecting cellulose microfibril orientation and pectin cross-linking strength, which together control wall plasticity and extensibility [92,93]. Spatially specific gradients of Ca²⁺ concentration have been observed in cell elongation zones, guiding polarized cell expansion and morphogenesis. Thus, calcium signaling functions not only as the chemical foundation for cell wall formation but also as a pivotal signaling hub for plant growth, development, and environmental adaptation. 

5.4. Regulatory mechanisms of calcium signaling in membrane stability maintenance and signal transduction 

Ca²⁺ plays a central regulatory role in maintaining the structural stability and functional integrity of plant cell membranes. As the interface between intracellular and extracellular environments, the cell membrane not only serves as a selective barrier for material exchange but also acts as a crucial platform for signal perception and transduction. Calcium interacts with membrane phospholipids, glycolipids, and membrane-associated proteins to form complex structures that alter surface charge distribution and lipid organization, thereby enhancing membrane stability and stress tolerance [94]. Under adverse conditions such as salinity, heat, or drought, membrane lipid peroxidation can lead to increased permeability; however, Ca²⁺ signaling mitigates ROS-induced oxidative damage by modulating lipid composition and activating antioxidant systems, including SOD and CAT, thus promoting membrane repair and homeostasis [95]. Moreover, Ca²⁺ regulates the activity of membrane-bound proteins such as plasma membrane H⁺-ATPases and ion transporters to maintain membrane potential and ionic balance, ensuring physiological stability under stress conditions [96]. 

Calcium signaling also plays a pivotal role in membrane-associated signal transduction. External stimuli such as mechanical perturbation, osmotic stress, or pathogen invasion can activate Ca²⁺-permeable channels in the plasma membrane, resulting in rapid Ca²⁺ influx and the formation of transient cytosolic “calcium waves” [97]. The spatiotemporal characteristics of these Ca²⁺ waves—including their amplitude, frequency, and duration—encode specific information that determines the type and intensity of the plant’s response. Effector molecules such as CaMs, CDPKs, and calcium/calmodulin-dependent protein kinases (CCaMKs) perceive cytosolic Ca²⁺ fluctuations and mediate downstream responses by regulating gene transcription, ion channel gating, and metabolic networks. For example, during pathogen attack, Ca²⁺ influx activates CDPK-mediated defense signaling cascades, leading to the induction of pathogenesis-related (PR) genes and strengthening of the plant immune response [98]. 

In addition, calcium signaling interacts with other pathways—including ROS, hormone, and nitric oxide signaling—to form a complex, multilayered regulatory network for membrane signal transmission. A bidirectional feedback mechanism exists between Ca²⁺ and ROS signaling: Ca²⁺ activates NADPH oxidases to enhance ROS production, while ROS in turn modulates Ca²⁺ channel activity, thereby amplifying the signaling effect [99]. Meanwhile, Ca²⁺ and ABA jointly regulate stomatal movement and membrane potential variation to coordinate water balance under drought stress [100]. Through this synergistic network, Ca²⁺ functions not only as an independent messenger in environmental responses but also as a central integrator, ensuring precise control of membrane signal transduction and systemic homeostasis. 

6. Application prospects of calcium signaling pathway optimization

6.1. Crop breeding potential of calcium signaling pathway optimization

The optimization of calcium signaling pathways exhibits broad application potential in modern crop breeding. Ca²⁺, as a crucial secondary messenger in plant cells, is deeply involved in regulating diverse physiological and metabolic processes, including cell division and elongation, tissue differentiation, stress response, and signal integration. Precise modulation of calcium signaling pathways can significantly enhance crop adaptability and yield stability under complex environmental conditions. The primary objective of calcium signaling optimization is to strengthen crop resistance to drought, salinity, extreme temperatures, and pathogenic stress, thereby improving physiological robustness and yield potential. For instance, under drought stress, calcium signaling modulates stomatal movement, activates antioxidant defense systems, and maintains cellular osmotic balance to enhance water use efficiency and reduce transpiration-driven water loss, ensuring growth and reproduction under adverse conditions.

With the rapid advancement of molecular breeding and precision agriculture, the optimization of calcium signaling pathways is transitioning from theoretical exploration to practical application. Gene-editing technologies such as CRISPR-Cas9 provide precise tools for modifying calcium signaling-related genes by targeting key regulators such as CDPKs, CaMs, and calcium channels. Enhancing CDPK-mediated signaling can increase plant tolerance to salinity and oxidative stress, while modulating CaM expression improves intracellular calcium sensitivity and feedback regulation. Concurrently, the integration of AI and multi-omics data analysis has made calcium signaling optimization more efficient and predictive. AI algorithms can mine vast genomic, transcriptomic, and phenotypic datasets to identify regulatory factors closely associated with calcium signaling, assisting researchers in defining optimal gene combinations and breeding routes, and achieving precise genotype–phenotype mapping.

Amid global climate change and the increasing scarcity of arable land, developing high-yield, stress-tolerant, and resource-efficient crops has become central to sustainable agricultural development. Calcium signaling pathway optimization breeding—integrating gene editing, AI-assisted selection, and high-throughput phenotyping—enables the rapid creation of resilient and high-performance crop varieties. This strategy not only reduces reliance on chemical inputs and promotes ecological and green agriculture but also provides new insights into the precise regulation of plant defense mechanisms. In the future, systematic optimization of calcium signaling pathways is expected to become a key direction in molecular design breeding, offering a solid scientific foundation for global food security and sustainable agricultural advancement.

6.2. Calcium supply strategies in precision agriculture

In precision agriculture systems, optimizing calcium supply is a critical factor in improving crop productivity and quality. As an essential macronutrient, calcium not only plays a vital role in maintaining cell wall stability and membrane integrity but also functions as a secondary messenger involved in multiple signaling and stress response processes. Ensuring an adequate and continuous calcium supply throughout the crop growth cycle is fundamental to promoting healthy development, enhancing stress resistance, and improving yield quality. Precision agriculture, supported by sensor networks, drone monitoring, and satellite remote sensing technologies, enables real-time monitoring of soil calcium content, crop nutritional status, and environmental factors. This allows for dynamic adjustment of calcium application based on crop growth stages, effectively preventing both overfertilization and calcium deficiency.

The integration of AI and big data analytics provides scientific decision support for precise calcium management. By comprehensively analyzing soil nutrient data, climatic parameters, crop genotypes, and growth models, AI algorithms can accurately predict calcium demand at different developmental stages and offer real-time application recommendations. This ensures adequate calcium supply during critical growth periods such as seedling establishment, flowering, and fruit expansion. Through AI-driven dynamic regulation mechanisms, calcium fertilization can be managed with precision according to real-time crop requirements, minimizing fertilizer waste and environmental impact while significantly enhancing crop quality. For example, sufficient calcium supply can improve fruit firmness, extend storage life, and effectively prevent calcium-deficiency-related physiological disorders such as blossom-end rot in tomatoes and bitter pit in apples.

Moreover, optimizing calcium supply strategies extends beyond fertilization management to include the improvement of calcium formulations and the diversification of application methods. The effectiveness of traditional calcium fertilizers is often limited by factors such as soil pH, rhizosphere conditions, and ion competition. In contrast, novel formulations—including slow-release and foliar-applied calcium fertilizers—can markedly enhance calcium use efficiency. With the aid of advanced sensing and imaging technologies, farmers can monitor leaf calcium content in real time and dynamically adjust the frequency and concentration of foliar calcium applications for optimal absorption. Additionally, integrating drip irrigation and fertigation systems enables direct delivery of calcium sources to the rhizosphere, further improving application precision and uptake efficiency. Through multi-level and multi-modal optimization of calcium supply, crop stability and productivity can be effectively increased, resource waste reduced, and technological support provided for the green and sustainable development of agriculture.

6.3. Interactions between calcium-related genes and environmental regulation

Calcium-related genes play a central role in the molecular regulation of plant responses to environmental changes and stress adaptation. These genes not only participate directly in the generation and transduction of calcium signals but also interact with external environmental cues to modulate plant adaptation to various biotic and abiotic stresses. Environmental fluctuations—such as changes in temperature, light, water availability, and soil nutrients—can trigger dynamic alterations in intracellular free Ca²⁺ concentration, thereby activating calcium signaling networks that induce gene expression and regulate diverse physiological and metabolic processes. The expression patterns of calcium-related genes are highly dynamic and environment-dependent, particularly under stress conditions such as drought, salinity, and pathogen attack. In such contexts, calcium signaling rapidly senses stress stimuli and initiates a cascade of defense responses, including stomatal closure, osmotic adjustment, and activation of antioxidant mechanisms, which collectively enhance the plant’s environmental resilience.

The interaction between calcium-related genes and environmental regulation is most evident in plants’ rapid responses to stress conditions. For instance, during drought stress, intracellular Ca²⁺ levels rise sharply, activating key signaling molecules such as CDPKs, CaM, and CaM-interacting proteins. These molecules regulate the expression of drought-responsive genes, promote stomatal closure to minimize water loss, enhance root water uptake, and increase the activity of ROS-scavenging systems, thereby mitigating oxidative damage caused by dehydration. Similarly, under salt stress, calcium signaling regulates the activity of calcium channels, transporters, and ion pumps to maintain ionic homeostasis and protect plants from high salinity. Moreover, calcium signaling forms synergistic networks with phytohormones such as ABA, JA, and ethylene (ET), achieving spatiotemporal coordination of stress responses and adaptive gene expression.

The integration of AI technologies has greatly advanced the understanding of interactions between calcium-related genes and environmental regulation. By integrating genomics, transcriptomics, and multidimensional environmental data, AI enables the efficient identification of calcium-related gene expression patterns under various stress conditions and elucidates their potential correlations with environmental stimuli. Machine learning and deep learning algorithms can extract critical features from large-scale datasets to construct calcium signaling–environment regulation network models, predicting key regulatory nodes and signal transduction pathways. This data-driven approach not only enhances the systematic and precise analysis of calcium-mediated signaling networks but also provides new insights into the mechanisms of environmental adaptation. AI-driven systems biology thus offers a theoretical foundation for breeding crops with high calcium efficiency and strong stress resistance, laying a technological basis for precise calcium signaling regulation and intelligent crop improvement.

7. Conclusions and prospects

(1) The core role of calcium signaling. Ca²⁺, as one of the most critical secondary messengers in plant cells, participates throughout the entire process of plant growth, development, and stress response. By modulating diverse physiological and metabolic processes, calcium signaling enables plants to rapidly perceive and adapt to environmental fluctuations, thereby maintaining cellular homeostasis and physiological balance. Through interactions with calmodulins (CaM), CDPKs, and downstream regulatory factors, Ca²⁺ translates external stimuli into specific molecular signals that regulate gene expression, ion homeostasis, and signal integration. Studies have shown that calcium signaling plays a fundamental role not only in plant growth regulation, cell division, and developmental processes but also in responses to drought, salinity, temperature stress, and pathogen invasion. It forms the central regulatory network of plant stress resistance, providing a theoretical foundation for elucidating stress tolerance mechanisms and improving crop resilience.

(2) Integration of omics technologies and artificial intelligence. With the rapid advancement of multi-omics technologies, research on calcium-related genes and signaling pathways has entered a systematic and data-driven era. The accumulation of large-scale genomic, transcriptomic, and proteomic data provides a robust foundation for elucidating calcium signaling mechanisms. However, the multidimensional complexity of these datasets presents significant challenges for traditional analytical approaches. The incorporation of AI technologies has created new opportunities for mining and integrating large-scale omics data. Using machine learning, deep learning, and GNNs, researchers can efficiently identify key regulatory genes in calcium signaling pathways, predict gene–gene interactions, and reconstruct the network topology of calcium signaling. Moreover, AI algorithms can extract latent regulatory modules and cross-pathway nodes from multidimensional data, supporting the development of dynamic calcium signaling models. This integrative strategy not only enhances analytical efficiency and predictive accuracy but also establishes a methodological basis for the functional interpretation and applied transformation of calcium signaling research.

(3) Application of calcium signaling pathways in crop breeding. Precise regulation of calcium signaling pathways holds significant potential for crop genetic improvement. By modulating the expression of calcium-dependent genes and their signal transduction pathways, it is possible to substantially enhance crop stress tolerance, environmental adaptability, and resource use efficiency. Under abiotic and biotic stresses such as drought, salinity, and pathogen attack, optimizing calcium signaling pathways can promote the establishment of more efficient physiological defense systems, thereby achieving stable yields and improved productivity. At present, the integration of gene-editing technologies (e.g., CRISPR-Cas9) with AI-driven gene screening strategies allows researchers to more accurately identify and modify key calcium signaling genes, reshaping calcium regulatory networks to develop new crop varieties with high yield, strong stress resistance, and superior quality. Furthermore, studies on the interactions between calcium signaling and other regulatory systems—such as phytohormone and redox signaling—provide new perspectives for multi-trait coordinated improvement.

(4) Future research challenges and directions. Despite remarkable progress, the complexity of calcium signaling pathways remains a major challenge for future research. The spatiotemporal characteristics of Ca²⁺ dynamics, pathway redundancy, and multilayered crosstalk with ROS and hormone signaling networks require deeper exploration through the integration of systems biology and multi-omics approaches. Future efforts should focus on developing quantitative and predictive models of calcium signaling dynamics to achieve an integrated understanding from molecular mechanisms to system-level regulation. Moreover, with the exponential growth of omics data, leveraging AI algorithms for efficient data integration, knowledge extraction, and translation into actionable breeding strategies will be a central direction. Bridging basic research with applied innovation will be essential—through precise, intelligent, and visualized calcium signaling regulation—to realize synergistic progress in stress-resilient crop breeding and precision agriculture, ultimately contributing to the development of sustainable agricultural systems.

References

[1] Kudla J, Becker D, Grill E, et al. Advances and current challenges in calcium signaling. New Phytologist, 2018, 218(2), 414-431. 

[2] Thor K. Calcium—nutrient and messenger. Frontiers in Plant Science, 2019, 10, 440. 

[3] Zhao C, Zhang H, Song C, et al. Mechanisms of plant responses and adaptation to soil salinity. The Innovation, 2020, 1(1), 100017. 

[4] Demidchik V, Shabala S, Isayenkov S, et al. Calcium transport across plant membranes: Mechanisms and functions. New Phytologist, 2018, 220(1), 49-69. 

[5] Luan S, Wang C. Calcium signaling mechanisms across kingdoms. Annual Review of Cell and Developmental Biology, 2021, 37, 311-340. 

[6] Boudsocq M, Sheen J. CDPKs in immune and stress signaling. Trends in Plant Science, 2013, 18(1), 30-40. 

[7] Zeng H, Xu L, Singh A, et al. Involvement of calmodulin and calmodulin-like proteins in plant responses to abiotic stresses. Frontiers in Plant Science, 2015, 6, 600. 

[8] Dodd A N, Kudla J, Sanders D. The language of calcium signaling. Annual Review of Plant Biology, 2010, 61, 593-620. 

[9] Steinhorst L, Kudla J. Signaling in cells and organisms—calcium holds the line. Current Opinion in Plant Biology, 2014, 22, 14-21. 

[10] Batistič O, Kudla J. Plant calcineurin B-like proteins and their interacting protein kinases. Biochimica et Biophysica Acta - Molecular Cell Research, 2009, 1793(6), 985-992. 

[11] Ma W, Noble W S, Bailey-Serres J. Reconstruction of cell type-specific gene regulatory networks during hypoxia response in Arabidopsis thaliana. Genome Biology, 2019, 20(1), 274. 

[12] Rhee S Y, Bieker M, Hewett Hazelton N. Towards community standards in the quest for orthologs. Briefings in Bioinformatics, 2017, 18(6), 968-979. 

[13] Misra B B, Langefeld C, Olivier M, et al. Integrated omics: tools, advances and future approaches. Journal of Molecular Endocrinology, 2019, 62(1), R21-R45. 

[14] Ma C, Zhang H H, Wang X. Machine learning for big data analytics in plants. Trends in Plant Science, 2014, 19(12), 798-808. 

[15] Shao J, Zhang Y, Yu J, et al. Isolation, characterization, and expression analysis of 46 calcium-dependent protein kinase genes in rice. Plant Molecular Biology Reporter, 2020, 38(2), 164-182. 

[16] Liang B, Sun Y, Wang J, et al. Tomato protein phosphatase 2C influences the onset of flowering. New Phytologist, 2021, 232(6), 2501-2515. 

[17] White P J. Calcium channels in the plasma membrane of root cells. Annals of Botany, 2013, 112(2), 369-383. 

[18] Ranty B, Aldon D, Cotelle V, et al. Calcium sensors as key hubs in plant responses to biotic and abiotic stresses. Frontiers in Plant Science, 2016, 7, 327. 

[19] Virdi A S, Singh S, Singh P. Abiotic stress responses in plants: roles of calmodulin-regulated proteins. Frontiers in Plant Science, 2015, 6, 809. 

[20] Feng W, Kita D, Peaucelle A, et al. The FERONIA receptor kinase maintains cell-wall integrity during salt stress through Ca2+ signaling. Current Biology, 2018, 28(5), 666-675. 

[21] Asano T, Hayashi N, Kobayashi M, et al. A rice calcium-dependent protein kinase OsCPK12 oppositely modulates salt-stress tolerance and blast disease resistance. The Plant Journal, 2012, 69(1), 26-36. 

[22] Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics, 2008, 9(1), 559. 

[23] Camacho D M, Collins K M, Powers R K, et al. Next-generation machine learning for biological networks. Cell, 2018, 173(7), 1581-1592. 

[24] Krassowski M, Das V, Sahu S K, et al. State of the field in multi-omics research: from computational needs to data mining and sharing. Frontiers in Genetics, 2020, 11, 610798. 

[25] Cheval C, Aldon D, Galaud J P, et al. Calcium/calmodulin-mediated regulation of plant immunity. Biochimica et Biophysica Acta - Molecular Cell Research, 2013, 1833(7), 1766-1771. 

[26] Schulz P, Herde M, Romeis T. Calcium-dependent protein kinases: hubs in plant stress signaling and development. Plant Physiology, 2013, 163(2), 523-530. 

[27] Batistič O, Kudla J. Analysis of calcium signaling pathways in plants. Biochimica et Biophysica Acta - General Subjects, 2012, 1820(8), 1283-1293. 

[28] Ye N, Zhu G, Liu Y, et al. ABA controls H2O2 accumulation through the induction of OsCATB in rice leaves under water stress. Plant and Cell Physiology, 2011, 52(4), 689-698. 

[29] Gong Z, Xiong L, Shi H, et al. Plant abiotic stress response and nutrient use efficiency. Science China Life Sciences, 2020, 63(5), 635-674. 

[30] Libbrecht M W, Noble W S. Machine learning applications in genetics and genomics. Nature Reviews Genetics, 2015, 16(6), 321-332. 

[31] Singh A, Ganapathysubramanian B, Singh A K, et al. Machine learning for high-throughput stress phenotyping in plants. Trends in Plant Science, 2016, 21(2), 110-124. 

[32] Eraslan G, Avsec Ž, Gagneur J, et al. Deep learning: new computational modelling techniques for genomics. Nature Reviews Genetics, 2019, 20(7), 389-403. 

[33] Tian W, Wang C, Gao Q, et al. Calcium spikes, waves and oscillations in plant development and biotic interactions. Nature Plants, 2020, 6(7), 750-759. 

[34] Kwaaitaal M, Huisman R, Maintz J, et al. Ionotropic glutamate receptor (iGluR)-like channels mediate MAMP-induced calcium influx in Arabidopsis thaliana. Biochemical Journal, 2011, 440(3), 355-373. 

[35] Shi S, Li S, Asim M, et al. The Arabidopsis calcium-dependent protein kinases (CDPKs) and their roles in plant growth regulation and abiotic stress responses. International Journal of Molecular Sciences, 2018, 19(7), 1900. 

[36] Kolukisaoglu Ü, Weinl S, Blazevic D, et al. Calcium sensors and their interacting protein kinases: genomics of the Arabidopsis and rice CBL-CIPK signaling networks. Plant Physiology, 2004, 134(1), 43-58. 

[37] Ding Y, Shi Y, Yang S. Advances and challenges in uncovering cold tolerance regulatory mechanisms in plants. New Phytologist, 2019, 222(4), 1690-1704. 

[38] Sperschneider J. Machine learning in plant-pathogen interactions: Empowering biological predictions from field scale to genome scale. New Phytologist, 2020, 228(1), 35-41. 

[39] Ching T, Himmelstein D S, Beaulieu-Jones B K, et al. Opportunities and obstacles for deep learning in biology and medicine. Journal of the Royal Society Interface, 2018, 15(141), 20170387. 

[40] DeFalco T A, Bender K W, Snedden W A. Breaking the code: Ca2+ sensors in plant signalling. Biochemical Journal, 2010, 425(1), 27-40. 

[41] Choi W G, Toyota M, Kim S H, et al. Salt stress-induced Ca2+ waves are associated with rapid, long-distance root-to-shoot signaling in plants. Proceedings of the National Academy of Sciences, 2014, 111(17), 6497-6502. 

[42] Marcec M J, Gilroy S, Poovaiah B W, et al. Mutual interplay of Ca2+ and ROS signaling in plant immune response. Plant Science, 2019, 283, 343-354. 

[43] Weinl S, Kudla J. The CBL-CIPK Ca2+-decoding signaling network: function and perspectives. New Phytologist, 2009, 184(3), 517-528. 

[44] Tang R J, Zhao F G, Garcia V J, et al. Tonoplast CBL-CIPK calcium signaling network regulates magnesium homeostasis in Arabidopsis. Proceedings of the National Academy of Sciences, 2015, 112(10), 3134-3139. 

[45] McAinsh M R, Pittman J K. Shaping the calcium signature. New Phytologist, 2009, 181(2), 275-294. 

[46] Hashimoto K, Kudla J. Calcium decoding mechanisms in plants. Biochimie, 2011, 93(12), 2054-2059. 

[47] Jiang Z, Zhou X, Tao M, et al. Plant cell-surface GIPC sphingolipids sense salt to trigger Ca2+ influx. Nature, 2019, 572(7769), 341-346. 

[48] Wang H, Kong C, Yao Y, et al. Machine learning of plasma metabolome identifies biomarker panels for metabolic syndrome: findings from the China Suboptimal Health Cohort. Cardiovascular Diabetology, 2022, 21(1), 280. 

[49] Angermueller C, Pärnamaa T, Parts L, et al. Deep learning for computational biology. Molecular Systems Biology, 2016, 12(7), 878. 

[50] Zhou J, Cui G, Hu S, et al. Graph neural networks: A review of methods and applications. AI Open, 2020, 1, 57-81. 

[51] Zitnik M, Agrawal M, Leskovec J. Modeling polypharmacy side effects with graph convolutional networks. Bioinformatics, 2018, 34(13), i457-i466. 

[52] Chen T, Guestrin C. XGBoost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016, 785-794. 

[53] Qi Y. Random forest for bioinformatics. Ensemble machine learning, 2012, 307-323. 

[54] Noble W S. What is a support vector machine? Nature Biotechnology, 2006, 24(12), 1565-1567. 

[55] Zou Q, Qu K, Luo Y, et al. Predicting diabetes mellitus with machine learning techniques. Frontiers in Genetics, 2018, 9, 515. 

[56] Monshausen G B, Bibikova T N, Weisenseel M H, et al. Ca2+ regulates reactive oxygen species production and pH during mechanosensing in Arabidopsis roots. The Plant Cell, 2009, 21(8), 2341-2356. 

[57] Toyota M, Spencer D, Sawai-Toyota S, et al. Glutamate triggers long-distance, calcium-based plant defense signaling. Science, 2018, 361(6407), 1112-1115. 

[58] Du L, Ali G S, Simons K A, et al. Ca2+/calmodulin regulates salicylic-acid-mediated plant immunity. Nature, 2009, 457(7233), 1154-1158. 

[59] Asai S, Ohta K, Yoshioka H. MAPK signaling regulates nitric oxide and NADPH oxidase-dependent oxidative bursts in Nicotiana benthamiana. The Plant Cell, 2008, 20(5), 1390-1406. 

[60] Mittler R, Vanderauwera S, Suzuki N, et al. ROS signaling: the new wave?. Trends in Plant Science, 2011, 16(6), 300-309. 

[61] Xu J, Zhang S. Mitogen-activated protein kinase cascades in signaling plant growth and development. Trends in Plant Science, 2015, 20(1), 56-64. 

[62] Plieth C. Calcium: just another regulator in the machinery of life?. Annals of Botany, 2016, 118(5), 919-923. 

[63] Wu Z, Pan S, Chen F, et al. A comprehensive survey on graph neural networks. IEEE Transactions on Neural Networks and Learning Systems, 2021, 32(1), 4-24. 

[64] Gaudelet T, Day B, Jamasb A R, et al. Utilizing graph machine learning within drug discovery and development. Briefings in Bioinformatics, 2021, 22(6), bbab159. 

[65] Velickovic P, Cucurull G, Casanova A, et al. Graph attention networks. International Conference on Learning Representations, 2018, 1-12. 

[66] Breiman L. Random forests. Machine Learning, 2001, 45(1), 5-32. 

[67] Zeng X, Song X, Ma T, et al. Repurpose open data to discover therapeutics for COVID-19 using deep learning. Journal of Proteome Research, 2020, 19(11), 4624-4636. 

[68] Serin E A, Nijveen H, Hilhorst H W, et al. Learning from co-expression networks: Possibilities and challenges. Frontiers in Plant Science, 2016, 7, 444. 

[69] Zhang B, Horvath S. A general framework for weighted gene co-expression network analysis. Statistical Applications in Genetics and Molecular Biology, 2005, 4(1), Article17. 

[70] Usadel B, Obayashi T, Mutwil M, et al. Co-expression tools for plant biology: Opportunities for hypothesis generation and caveats. Plant, Cell & Environment, 2009, 32(12), 1633-1651. 

[71] Rhee S Y, Mutwil M. Towards revealing the functions of all genes in plants. Trends in Plant Science, 2014, 19(4), 212-221. 

[72] Movahedi A, Zhang J, Amirian R, et al. An overview of systems biology. Methods in Molecular Biology, 2018, 1702, 1-37. 

[73] Alipanahi B, Delong A, Weirauch M T, et al. Predicting the sequence specificities of DNA- and RNA-binding proteins by deep learning. Nature Biotechnology, 2015, 33(8), 831-838. 

[74] Harper J F, Breton G, Harmon A. Decoding Ca2+ signals through plant protein kinases. Annual Review of Plant Biology, 2004, 55, 263-288. 

[75] Franz S, Ehlert B, Liese A, et al. Calcium-dependent protein kinase CPK13 functions in guttation fluid secretion in Arabidopsis. New Phytologist, 2011, 189(1), 94-106. 

[76] Zhao R, Sun H L, Mei C, et al. The Arabidopsis Ca2+-dependent protein kinase CPK12 negatively regulates abscisic acid signaling in seed germination and post-germination growth. New Phytologist, 2011, 192(1), 61-73. 

[77] Swan A L, Mobasheri A, Allaway D, et al. Application of machine learning to proteomics data: classification and biomarker identification in postgenomics biology. OMICS: A Journal of Integrative Biology, 2013, 17(12), 595-610. 

[78] Greener J G, Kandathil S M, Moffat L, et al. A guide to machine learning for biologists. Nature Reviews Molecular Cell Biology, 2022, 23(1), 40-55. 

[79] Hepler P K. Calcium: a central regulator of plant growth and development. The Plant Cell, 2005, 17(8), 2142-2155. 

[80] Zou J J, Wei F J, Wang C, et al. Arabidopsis calcium-dependent protein kinase CPK10 functions in abscisic acid- and Ca2+-mediated stomatal regulation in response to drought stress. Plant Physiology, 2010, 154(3), 1232-1243. 

[81] Wilkinson S, Davies W J. Drought, ozone, ABA and ethylene: New insights from cell to plant to community. Plant, Cell & Environment, 2010, 33(4), 510-525. 

[82] Sharma A, Shahzad B, Rehman A, et al. Response of phenylpropanoid pathway and the role of polyphenols in plants under abiotic stress. Molecules, 2019, 24(13), 2452. 

[83] Tohge T, de Souza L P, Fernie A R. Current understanding of the pathways of flavonoid biosynthesis in model and crop plants. Journal of Experimental Botany, 2017, 68(15), 4013-4028. 

[84] Fiehn O. Metabolomics—the link between genotypes and phenotypes. Plant Molecular Biology, 2002, 48(1-2), 155-171. 

[85] Liebal U W, Phan A N T, Sudhakar M, et al. Machine learning applications for mass spectrometry-based metabolomics. Metabolites, 2020, 10(6), 243. 

[86] Witting M, Böcker S. Current status of retention time prediction in metabolite identification. Journal of Separation Science, 2020, 43(9-10), 1746-1754. 

[87] Weckwerth W. Metabolomics: an integral technique in systems biology. Bioanalysis, 2010, 2(4), 829-836. 

[88] Redestig H, Costa I G. Detection and interpretation of metabolite-transcript coresponses using combined profiling data. Bioinformatics, 2011, 27(13), i357-i365. 

[89] Peaucelle A, Braybrook S A, Le Guillou L, et al. Pectin-induced changes in cell wall mechanics underlie organ initiation in Arabidopsis. Current Biology, 2011, 21(20), 1720-1726. 

[90] Wolf S, Hématy K, Höfte H. Growth control and cell wall signaling in plants. Annual Review of Plant Biology, 2012, 63, 381-407. 

[91] Denness L, McKenna J F, Segonzac C, et al. Cell wall damage-induced lignin biosynthesis is regulated by a reactive oxygen species- and jasmonic acid-dependent process in Arabidopsis. Plant Physiology, 2011, 156(3), 1364-1374. 

[92] Cosgrove D J. Growth of the plant cell wall. Nature Reviews Molecular Cell Biology, 2005, 6(11), 850-861. 

[93] Hocq L, Pelloux J, Lefebvre V. Connecting homogalacturonan-type pectin remodeling to acid growth. Trends in Plant Science, 2017, 22(1), 20-29. 

[94] Haswell E S, Verslues P E. The ongoing search for the molecular basis of plant osmosensing. Journal of General Physiology, 2015, 145(5), 389-394. 

[95] Apel K, Hirt H. Reactive oxygen species: metabolism, oxidative stress, and signal transduction. Annual Review of Plant Biology, 2004, 55, 373-399. 

[96] Palmgren M G, Nissen P. P-type ATPases. Annual Review of Biophysics, 2011, 40, 243-266. 

[97] Behera S, Zhaolong X, Luoni L, et al. Cellular Ca2+ signals generate defined pH signatures in plants. The Plant Cell, 2018, 30(11), 2704-2719. 

[98] Gao X, Chen X, Lin W, et al. Bifurcation of Arabidopsis NLR immune signaling via Ca2+-dependent protein kinases. PLoS Pathogens, 2013, 9(1), e1003127. 

[99] Kaur G, Pati P K. Analysis of cis-acting regulatory elements of Respiratory burst oxidase homolog (Rboh) gene families in Arabidopsis and rice provides clues for their diverse functions. Computational Biology and Chemistry, 2016, 62, 104-118. 

[100] Pantin F, Monnet F, Jannaud D, et al. The dual effect of abscisic acid on stomata. New Phytologist, 2013, 197(1), 65-72.