抗病耐逆基因鉴定、智能设计育种

Potato is the world's fourth largest food crop yet it continues to endure late blight, a devastating disease caused by the Irish famine pathogen Phytophthora infestans. Breeding broad-spectrum disease resistance (R) genes into potato (Solanum tuberosum) is the best strategy for genetically managing late blight but current approaches are slow and inefficient. We used a repertoire of effector genes predicted computationally from the P. infestans genome to accelerate the identification, functional characterization, and cloning of potentially broad-spectrum R genes. An initial set of 54 effectors containing a signal peptide and a RXLR motif was profiled for activation of innate immunity (avirulence or Avr activity) on wild Solanum species and tentative Avr candidates were identified. The RXLR effector family IpiO induced hypersensitive responses (HR) in S. stoloniferum, S. papita and the more distantly related S. bulbocastanum, the source of the R gene Rpi-blb1. Genetic studies with S. stoloniferum showed cosegregation of resistance to P. infestans and response to IpiO. Transient co-expression of IpiO with Rpi-blb1 in a heterologous Nicotiana benthamiana system identified IpiO as Avr-blb1. A candidate gene approach led to the rapid cloning of S. stoloniferum Rpi-sto1 and S. papita Rpi-pta1, which are functionally equivalent to Rpi-blb1. Our findings indicate that effector genomics enables discovery and functional profiling of late blight R genes and Avr genes at an unprecedented rate and promises to accelerate the engineering of late blight resistant potato varieties.

Summary Whole‐genome resequencing (WGRS) of mapping populations has facilitated development of high‐density genetic maps essential for fine mapping and candidate gene discovery for traits of interest in crop species. Leaf spots, including early leaf spot (ELS) and late leaf spot (LLS), and Tomato spotted wilt virus (TSWV) are devastating diseases in peanut causing significant yield loss. We generated WGRS data on a recombinant inbred line population, developed a SNP‐based high‐density genetic map, and conducted fine mapping, candidate gene discovery and marker validation for ELS, LLS and TSWV. The first sequence‐based high‐density map was constructed with 8869 SNPs assigned to 20 linkage groups, representing 20 chromosomes, for the ‘T’ population (Tifrunner × GT‐C20) with a map length of 3120 cM and an average distance of 1.45 cM. The quantitative trait locus (QTL) analysis using high‐density genetic map and multiple season phenotyping data identified 35 main‐effect QTLs with phenotypic variation explained (PVE) from 6.32% to 47.63%. Among major‐effect QTLs mapped, there were two QTLs for ELS on B05 with 47.42% PVE and B03 with 47.38% PVE, two QTLs for LLS on A05 with 47.63% and B03 with 34.03% PVE and one QTL for TSWV on B09 with 40.71% PVE. The epistasis and environment interaction analyses identified significant environmental effects on these traits. The identified QTL regions had disease resistance genes including R‐genes and transcription factors. KASP markers were developed for major QTLs and validated in the population and are ready for further deployment in genomics‐assisted breeding in peanut.

Next-generation sequencing technologies are accelerating gene discovery by combining multiple steps of mapping and cloning used in the traditional map-based approach into one step using DNA sequence polymorphisms existing between two different accessions/strains/backgrounds of the same species. The existing next-generation sequencing method, like the traditional one, requires the use of a segregating population from a cross of a mutant organism in one accession with a wild-type (WT) organism in a different accession. It therefore could potentially be limited by modification of mutant phenotypes in different accessions and/or by the lengthy process required to construct a particular mapping parent in a second accession. Here we present mapping and cloning of an enhancer mutation with next-generation sequencing on bulked segregants in the same accession using sequence polymorphisms induced by a chemical mutagen. This method complements the conventional cloning approach and makes forward genetics more feasible and powerful in molecularly dissecting biological processes in any organisms. The pipeline developed in this study can be used to clone causal genes in background of single mutants or higher order of mutants and in species with or without sequence information on multiple accessions.

… conferred resistance (Yorinori et al. 2005). Therefore, the discovery of new resistance genes is … the possibility that two independent genes are involved in disease resistance in these PIs. …

Pathogens are among the most limiting factors for crop success and expansion. Thus, finding the underlying genetic cause of pathogen resistance is the main goal for plant geneticists. The activation of a plant’s immune system is mediated by the presence of specific receptors known as disease-resistance genes (R genes). Typical R genes encode functional immune receptors with nucleotide-binding sites (NBS) and leucine-rich repeat (LRR) domains, making the NBS-LRRs the largest family of plant resistance genes. Establishing host resistance is crucial for plant growth and crop yield but also for reducing pesticide use. In this regard, pyramiding R genes is thought to be the most ecologically friendly way to enhance the durability of resistance. To accomplish this, researchers must first identify the related genes, or linked markers, within the genomes. However, the duplicated nature, with the presence of frequent paralogues, and clustered characteristic of NLRs make them difficult to predict with the classic automatic gene annotation pipelines. In the last several years, efforts have been made to develop new methods leading to a proliferation of reports on cloned genes. Herein, we review the bioinformatic tools to assist the discovery of R genes in plants, focusing on well-established pipelines with an important computer-based component.

Soybean mosaic virus (SMV) is a prevalent virus infecting soybean (Glycine max L. Merr) worldwide. The incorporation of Rsv4, conferring resistance to all currently known strains in the United States, can assist in creating durable virus resistance in soybean. Additionally, lines heterozygous at the Rsv4 locus often express a late susceptible phenotype, showing symptoms only in mid to late vegetative growth. In this study the whole‐genome shotgun sequence (WGS) of soybean was utilized for fine mapping and examining potential Rsv4 gene candidates in two populations. Six markers, designed from the WGS, were used to localize Rsv4 in the same, 1.3‐cM region in both mapping populations, a physical interval of less than 100 kb on chromosome 2. This region contained no sequences previously related to virus resistance, namely nucleotide binding site‐leucine rich repeat gene sequences or eukaryotic translation initiation factors. Instead, sequence analysis revealed several predicted transcription factors and unknown protein products. We conclude that Rsv4 likely belongs to a new class of resistance genes that interfere with viral infection and cell‐to‐cell movement, and delay vascular movement.

… Discovery of the structure of R genes and R gene loci provides insight into R … the biochemical and genetic basis of disease resistance (27, 64, 73), while the use of resistant cultivars has …

Recent work has shown that differentially expressed cDNA fragments identified during biotic challenge have great potential as genetic markers for pest and disease resistance in sugarcane. Responses to the smut fungus ( Ustilago scitaminea ), the stalk borer eldana ( Eldana saccharina ) and Sugarcane Mosaic Virus (SCMV) have been the main functional targets to date. Potentially useful cDNAs were identified using SSH and cDNA-AFLP differential display. These were isolated, sequenced and assigned putative functions by comparison with international databases. Using an RFLP approach, fragments with identities suggesting a role in resistance mechanisms were used as probes on a population of 78 sugarcane genotypes used in the breeding programme and well characterised for the traits of interest. Polymorphic markers were scored, and association with phenotype analysed using statistical methods developed in house. A set of 51 probes used for RFLP analysis has yielded 275 polymorphisms to date. Preliminary analyses of the data have identified 69 polymorphisms showing correlation with eldana resistance, 59 with smut, and 35 with SCMV. Most of the probes (76%) yielded at least one RFLP marker associated with smut, eldana or SCMV resistance, illustrating the efficiency of this marker generating strategy. Groups of uncorrelated markers have been found to be associated with each of the resistance traits under investigation, indicating the likelihood that each one represents a different aspect or mechanism of resistance. Markers identified in this project are already being used to devise crosses for the coordinated assembly of resistance factors in progeny. A subset of the most significant markers has been screened against a further collection of 53 varieties in order to extend the capacity of marker-assisted breeding. In addition, newly identified cDNAs are continually being investigated via RFLP screening.

… resistance, review technologies available for gene discovery, mechanisms of resistance, improvement of the genetic … disease resistance genes, such as in cassava against brown streak …

… of disease resistance forms the subject of this review and is known variously as race-specific resistance, gene-for-gene resistance… -threonine kinases, including the discovery of Fen, Ptil, …

Plant breeders have used disease resistance genes (R genes) to control plant disease since the turn of the century. Molecular cloning of R genes that enable plants to resist a diverse range of pathogens has revealed that the proteins encoded by these genes have several features in common. These findings suggest that plants may have evolved common signal transduction mechanisms for the expression of resistance to a wide range of unrelated pathogens. Characterization of the molecular signals involved in pathogen recognition and of the molecular events that specify the expression of resistance may lead to novel strategies for plant disease control.

… fundamental to plant development, metabolism and disease resistance in Arabidopsis. This review provides selected examples of these discoveries to highlight the utility of this …

… resistance in plant. As PRRs are able to perceive a great range of microbial elicitors, the discovery … to exploit the wide diversity of elicitor perception to enhance crop disease resistance. …

Gene discovery and marker development using DNA-based tools require plant populations with well-documented phenotypes. If dissimilar phenotype evaluation methods or data scoring techniques are used with different crops, or at different laboratories for the same crops, then data mining for genetic marker correlations is challenging. For example, apples and pears may share many of the same disease resistance genes. Fire blight resistance evaluations for apples often use a scale of 1 to 5 and pear evaluations use a scale of 1 to 9. In some reports, a low number means low susceptibility and in other reports, a low number means low resistance. Other disease evaluations rate resistance as greater than or less than a well-documented standard cultivar. Environment, pathogen isolate, and whether disease ratings are the result of natural infection or artificial inoculation also have a strong impact on disease resistance ratings. Before a wider set of disease resistance phenotype data can be correlated with genetic data, rating scales must be standardized and the evaluation environment must be taken into account. Standardizing the recording of disease resistance data in plant phenotype databases will improve the ability to correlate these data with genomic data.

… Although NBS sequences related to plant R-genes are being frequently discovered as part of genome initiatives and from ampli®cation using PCR with degenerate primers, it is not …

BackgroundDue to its origin, peanut has a very narrow genetic background. Wild relatives can be a source of genetic variability for cultivated peanut. In this study, the transcriptome of the wild species Arachis stenosperma accession V10309 was analyzed.ResultsESTs were produced from four cDNA libraries of RNAs extracted from leaves and roots of A. stenosperma. Randomly selected cDNA clones were sequenced to generate 8,785 ESTs, of which 6,264 (71.3%) had high quality, with 3,500 clusters: 963 contigs and 2537 singlets. Only 55.9% matched homologous sequences of known genes. ESTs were classified into 23 different categories according to putative protein functions. Numerous sequences related to disease resistance, drought tolerance and human health were identified. Two hundred and six microsatellites were found and markers have been developed for 188 of these. The microsatellite profile was analyzed and compared to other transcribed and genomic sequence data.ConclusionThis is, to date, the first report on the analysis of transcriptome of a wild relative of peanut. The ESTs produced in this study are a valuable resource for gene discovery, the characterization of new wild alleles, and for marker development. The ESTs were released in the [GenBank:EH041934 to EH048197].

… disease resistance genes (R genes) encode proteins that detect pathogens. R genes have been used in resistance … will enhance the use of R genes for disease control. Definition of …

… Building on this progress, the present study aims to (I) identify key genes and molecular … ; (II) identify drought-adapted wheat lines for breeding applications; and (III) explore genetic …

Agriculture faces significant obstacles in the form of unpredictable weather patterns. Artificial intelligence (AI) has emerged as a potential resource to mitigate this uncertainty. AI has significantly impacted the fields of crop breeding and plant science research, with the primary objective of ensuring global food security. AI empowers farmers to maintain ongoing surveillance of their crops and soil health, allowing them to promptly detect any crop stress and refine their agricultural methods accordingly. Pioneering AI technologies, including hyperspectral imaging and non-imaging sensors, provide nondestructive methods for prompt crop health evaluations. The integration of AI with phenomics and genomics data has accelerated the development of robust crop varieties. These varieties demonstrate significant yield potential and exhibit enhanced resilience to climate change. AI processes complex omics datasets, enabling the implementation of speed breeding procedures. Utilizing next-generation AI technology, such as artificial neural networks, genetic algorithms (GAs), and grey system theory (GST), has significant implications for enhancing resource utilization efficiency. Additionally, integrating AI with omics data derived from genome sequencing enables the development of crops resistant to climate challenges, thereby ensuring high yields while maintaining ecosystem preservation. These innovations play a crucial role in cultivating adaptable, high-yielding crops that are essential for feeding the world's growing population.

With the development of new technologies in recent years, researchers have made significant progress in crop breeding. Modern breeding differs from traditional breeding because of great changes in technical means and breeding concepts. Whereas traditional breeding initially focused on high yields, modern breeding focuses on breeding orientations based on different crops’ audiences or by-products. The process of modern breeding starts from the creation of material populations, which can be constructed by natural mutagenesis, chemical mutagenesis, physical mutagenesis transfer DNA (T-DNA), Tos17 (endogenous retrotransposon), etc. Then, gene function can be mined through QTL mapping, Bulked-segregant analysis (BSA), Genome-wide association studies (GWASs), RNA interference (RNAi), and gene editing. Then, at the transcriptional, post-transcriptional, and translational levels, the functions of genes are described in terms of post-translational aspects. This article mainly discusses the application of the above modern scientific and technological methods of breeding and the advantages and limitations of crop breeding and diversity. In particular, the development of gene editing technology has contributed to modern breeding research.

… of crop breeding data from artificial selection to intelligent design breeding. It explores the applications and development trends of AI and big data in modern crop breeding from several …

ABSTRACT The past decade has witnessed rapid developments in gene discovery, biological big data (BBD), artificial intelligence (AI)‐aided technologies, and molecular breeding. These advancements are expected to accelerate crop breeding under the pressure of increasing demands for food. Here, we first summarize current breeding methods and discuss the need for new ways to support breeding efforts. Then, we review how to combine BBD and AI technologies for genetic dissection, exploring functional genes, predicting regulatory elements and functional domains, and phenotypic prediction. Finally, we propose the concept of intelligent precision design breeding (IPDB) driven by AI technology and offer ideas about how to implement IPDB. We hope that IPDB will enhance the predictability, efficiency, and cost of crop breeding compared with current technologies. As an example of IPDB, we explore the possibilities offered by CropGPT, which combines biological techniques, bioinformatics, and breeding art from breeders, and presents an open, shareable, and cooperative breeding system. IPDB provides integrated services and communication platforms for biologists, bioinformatics experts, germplasm resource specialists, breeders, dealers, and farmers, and should be well suited for future breeding.

Recently, Artificial intelligence (AI) has emerged as a revolutionary field, providing a great opportunity in shaping modern crop breeding, and is extensively used indoors for plant science. Advances in crop phenomics, enviromics, together with the other "omics" approaches are paving ways for elucidating the detailed complex biological mechanisms that motivate crop functions in response to environmental trepidations. These "omics" approaches have provided plant researchers with precise tools to evaluate the important agronomic traits for larger-sized germplasm at a reduced time interval in the early growth stages. However, the big data and the complex relationships within impede the understanding of the complex mechanisms behind genes driving the agronomic-trait formations. AI brings huge computational power and many new tools and strategies for future breeding. The present review will encompass how applications of AI technology, utilized for current breeding practice, assist to solve the problem in high-throughput phenotyping and gene functional analysis, and how advances in AI technologies bring new opportunities for future breeding, to make envirotyping data widely utilized in breeding. Furthermore, in the current breeding methods, linking genotype to phenotype remains a massive challenge and impedes the optimal application of high-throughput field phenotyping, genomics, and enviromics. In this review, we elaborate on how AI will be the preferred tool to increase the accuracy in high-throughput crop phenotyping, genotyping, and envirotyping data; moreover, we explore the developing approaches and challenges for multiomics big computing data integration. Therefore, the integration of AI with "omics" tools can allow rapid gene identification and eventually accelerate crop-improvement programs.

… Artificial Intelligence (AI) into crop breeding represents a paradigm shift toward data-driven agricultural practices, aiming to enhance the efficiency and precision of crop … on crop breeding…

In climate change, breeding crop plants with improved productivity, sustainability, and adaptability has become a daunting challenge to ensure global food security for the ever-growing global population. Correspondingly, climate-smart crops are also the need to regulate biomass production, which is imperative for the maintenance of ecosystem services worldwide. Since conventional breeding technologies for crop improvement are limited, time-consuming, and involve laborious selection processes to foster new and improved crop varieties. An urgent need is to accelerate the plant breeding cycle using artificial intelligence (AI) to depict plant responses to environmental perturbations in real-time. The review is a collection of authorized information from various sources such as journals, books, book chapters, technical bulletins, conference papers, and verified online contents. Speed breeding has emerged as an essential strategy for accelerating the breeding cycles of crop plants by growing them under artificial light and temperature conditions. Furthermore, speed breeding can also integrate marker-assisted selection and cutting-edged gene-editing tools for early selection and manipulation of essential crops with superior agronomic traits. Scientists have recently applied next-generation AI to delve deeper into the complex biological and molecular mechanisms that govern plant functions under environmental cues. In addition, AIs can integrate, assimilate, and analyze complex OMICS data sets, an essential prerequisite for successful speed breeding protocol implementation to breed crop plants with superior yield and adaptability.

… crop improvement techniques can bridge the gap for feeding the ever-increasing population. Artificial intelligence (AI) … Precise phenotyping is a prerequisite to advance crop breeding for …

… underlying reasons, thereby making AI more closely resemble real… An ML ecosystem specifically designed for AI breeding in … will make plant breeding smarter and easier in this era of …

Applications of artificial intelligence, machine learning, and deep learning in plant breedingIn recent years, the field of plant breeding has witnessed a paradigm shift driven by advancements in artificial intelligence (AI) technologies, including machine learning (ML) and deep learning (DL) technologies.These cutting-edge techniques have transformed our understanding of plant biology.From decoding the intricate molecular mechanisms of plant defense to automating disease detection and optimizing nutrient levels, AI is reshaping the landscape of plant breeding (Hamazaki and Iwata, 2024).AI-assisted omics techniques offer insights into plant-pathogen interactions and facilitate the identification of stress-responsive genes (Mahmood et al., 2022;Chao et al., 2023).This Research Topic presents 16 papers on the application of computer techniques in plant science.Murmu et al. highlighted the potential of AI algorithms, particularly ML and DL, in decoding complex omics data to elucidate the molecular foundations of plant defense.In their review article, they explored AI-assisted omics techniques' applications, challenges, and prospects in enhancing crop protection strategies and ensuring global food security amidst environmental challenges.By integrating AI with omics technologies, researchers can unravel intricate gene regulatory networks and develop targeted interventions for enhancing crop resilience.As we confront the challenges of climate change and emerging diseases, AI-driven approaches offer a robust toolkit for ensuring global food security and sustainability in agriculture.Climate change poses significant threats to agricultural systems, emphasizing the importance of elucidating cold defense mechanisms in crops.Konecny et al. introduced the Self Organizing Maps (SOM)-based ML method to decipher gene expression patterns in Frontiers in Plant Science frontiersin.

The convergence of artificial intelligence (AI) and multi-omics technologies is revolutionizing crop breeding, offering solutions to pressing agricultural issues like climate change, food security, and crop resilience. Omics technologies encompassing genomics, transcriptomics, proteomics, metabolomics, and phenomics have transformed crop breeding by identifying genes, proteins, and metabolic pathways associated with important agronomic traits, including stress tolerance, yield, and nutritional content. However, the vast and intricate nature of multi-omics data necessitates sophisticated computational methods for dissecting these large-scale datasets. These AI-driven approaches enhance genome-wide association studies, quantitative trait loci (QTL) mapping, and genomic selection by establishing precise connections between genotype and phenotype. AI models can also predict gene functions, analyze extensive phenotypic data, and refine breeding strategies to produce climate-resilient crops. Furthermore, AI improves the synergy between different omics platforms, allowing for a deeper understanding of crop responses to both abiotic and biotic stressors. This chapter outlines recent progress in sequencing technologies and bioinformatics, which are speeding up the process of gene discovery and functional characterization. It underscores the importance of AI-assisted omics in linking environmental challenges to food security through molecular breeding and genome editing. By promoting collaboration between plant scientists, computational biologists, and agricultural experts, the integration of AI and omics provides a promising path toward developing resilient crops that can sustainably meet global food demands.

Clustered regularly interspaced short palindromic repeats (CRISPR)-based genome editing and artificial intelligence (AI) are increasingly presented as a unified route to precision plant breeding. Their convergence is scientifically plausible but unevenly demonstrated. CRISPR systems can create targeted sequence changes, whereas AI can prioritise candidate genes, integrate genomic and phenomic data, optimise guide RNAs and editors, predict editing outcomes, and support iterative genotype-to-phenotype learning. This critical narrative review evaluates the evidence linking these capabilities across the crop-improvement pipeline. Literature published from 1 January 2012 to 5 June 2026 was selected through transparent searches of accessible scholarly indexes and bibliographic resources, followed by citation tracking, metadata verification and thematic appraisal. Evidence is strongest for CRISPR-mediated improvement of discrete, biologically well-characterised traits, including disease resistance, quality attributes, plant architecture and selected stress responses. AI has also achieved useful performance in phenotyping, genomic prediction and CRISPR design, but superiority over conventional statistical or rule-based approaches is not consistent across datasets, species or prediction tasks. Direct evidence for fully integrated, AI-guided CRISPR breeding programmes that deliver stable field performance remains limited. Major constraints include uncertain causal target identification, small and non-representative training datasets, poor transferability across genetic backgrounds, polyploidy, genotype-by-environment interaction, transformation and regeneration bottlenecks, incomplete detection of unintended outcomes, and heterogeneous regulation. The most defensible interpretation is therefore that AI and CRISPR are complementary components of an emerging design-build-test-learn framework rather than a mature autonomous breeding platform. Progress will depend on plant-specific benchmark datasets, prospective validation, multi-environment field trials, interoperable data standards, equitable access to transformation and computational infrastructure, and governance focused on the properties and evidence of resulting products. Their integration can accelerate precision breeding, but biological causality, experimental validation and breeding judgement remain indispensable.

Plant breeding is crucial for addressing global challenges like food security, climate change resilience, and sustainable agriculture. The integration of Artificial Intelligence (AI) and Machine Learning (ML) techniques has revolutionized traditional breeding methods, enabling the development of improved crop varieties. AI and ML algorithms are used for tasks such as genotype-phenotype prediction, genomic selection, trait discovery, and optimization of breeding schemes. These technologies help identify genetic markers associated with desirable traits, enabling breeders to select plants with desired characteristics more efficiently. AI-driven models can predict the performance of novel genotypes under different environmental conditions, aiding in the development of resilient and high-yielding crop varieties. AI-powered tools can optimize breeding strategies by simulating breeding outcomes, reducing time and resource constraints. However, challenges such as data quality, model interpretability, and ethical considerations need to be addressed. Additionally, the accessibility of advanced computational resources and expertise remains a barrier for many breeders, especially in developing countries. The future of AI and ML in plant breeding holds great promise, with continued advancements in computational biology, genomics, and data analytics. Collaboration between breeders, data scientists, and biotechnologists is essential for leveraging AI and ML technologies to their full potential in addressing global agricultural challenges.

[Significance]The escalating complexity of genotype-phenotype-environment interactions and the explosive growth of multi-omics big data have necessitated a paradigm shift in crop breeding from empirical selection to intelligent design (Breeding 5.0). The aim of this paper is to systematically explore the underlying logic of the AI-driven crop breeding paradigm shift, comprehensively outline its generational evolution, analyze its core technical implementations in phenomics, genomics, multi-omics integration, and molecular design, and dissect how large agricultural foundation models reconstruct the entire seed industry workflow.[Progress]The historical evolution of crop breeding was first traced from 1.0 empirical domestication to 5.0 smart Breeding characterized by the deep integration of biotechnology (BT) and information technology (IT). In germplasm resource evaluation, deep learning algorithms enabled high-dimensional pattern recognition and unsupervised feature compression to unlock rare alleles from unannotated sequences, large language models (LLMs) like PlantConnectome and wheat germplasm information extraction (WGIE) leverage retrieval-augmented generation (RAG) to automatically construct structural knowledge graphs from unstructured historical literature, achieving predictive and dynamic germplasm evaluations. In high-throughput phenotyping, industrial platforms capture 3D point cloud and multi-spectral data at 0.1 mm resolution, while convolutional neural networks couple with the integrated genomic-enviromic prediction (iGEP) framework to build full-lifecycle digital twin crop models in virtual space. Regarding genomic prediction, the limitations of linear paradigms were dissected and cutting-edge deep learning architectures were highlighted: SoyDNGP applied a 3D-CNN to map chromosomal topology for complex soybean traits; DPCformer employed self-attention mechanisms to dynamically calculate environmental weights under multi-adversarial constraints; and HyenaDNA utilized long-convolution filters to bypass the O(N2) computational complexity limitation of standard Transformers, reducing it to O(N·log N) for chromosome-scale modeling. For multi-omics integration, intermediate fusion strategies were elucidated for their superior capacity to capture cross-layer biological compensatory pathways. In molecular design breeding, foundational plant language models were highlighted, such as AgroNT for zero-shot expression prediction and OpenCRISPR-1, the world's first de novo AI-generated genome editor built to capture underlying physico-chemical syntax. Furthermore, the emergence logic of major domestic and international agricultural foundation models was analyzed through the mathematical lenses of scaling laws, parameter-efficient fine-tuning, and multi-task evaluation benchmarks. Finally, empirical effectiveness was comprehensively evaluated through multinational success stories, including Bayer's Climate FieldView, IRRI's night-temperature thermal models for "Green Super Rice", and China's state-led breeding platforms for stress-resistant maize and high-yield soybean.[Conclusions and Prospects]Key structural challenges that smart breeding faces are thoroughly dissected: data silos and standardization dilemmas, extreme computational resource asymmetry and high training costs, the lack of biological causal logic in deep learning "black boxes", and the structural scarcity of interdisciplinary BT-IT talent. To overcome these bottlenecks, future research and policy efforts should focus on four pillars: 1) Establishing standardized open data ecosystems following FAIR (Findable, Accessible, Interoperable, Reusable) principles and promoting federated learning under a national AI data copyright integration platform to safeguard digital borders; 2) Exploring cloud-to-edge lightweight model deployment pathways through parameter-efficient fine-tuning, quantization, and knowledge distillation to empower real-time field-side decision-making and realize compute equity; 3) Deepening the mechanistic integration of AI and synthetic biology by absorbing breakthroughs from multi-scale modeling frameworks into the virtual "design-build-test-learn" cycle to break natural evolutionary thresholds; and 4) improving regulatory approval frameworks for AI-designed crops and modernizing agricultural education to cultivate a new generation of digital agronomists.

Global food security is increasingly threatened by climate change, which intensifies both biotic and abiotic stresses on crops, leading to reduced yield stability and productivity. Conventional breeding approaches are insufficient to address these challenges due to long breeding cycles, strong genotype–environment interactions, and limited capacity to predict crop performance under future climate scenarios. Predictive artificial intelligence (AI) offers a powerful solution by integrating genomics, phenomics, environmental, and climate data to model complex traits, stress tolerance, and genotype performance across diverse agroecological conditions. This review synthesized recent advances in Machine Learning (ML), Deep Learning (DL), genomic prediction, high-throughput phenotyping, and climate modeling that collectively support the development of climate-resilient crop varieties. The application of predictive AI in plant breeding enhances selection accuracy, accelerates breeding decisions, reduces dependency on costly and time-consuming field trials, as well as improves resource-use efficiency. By enabling data-driven decision-making, AI-based approaches significantly improve the precision and scalability of modern crop improvement programs. Looking ahead, the integration of predictive AI with explainable modeling frameworks, multi-omics datasets, and genome-editing technologies holds significant promise for accelerating the development of high-yielding, resilient, and sustainable crops, thereby contributing to long-term food security under changing climatic conditions.

Plant Breeding is a reliable assurance of agriculture productivity and nutritional security, while information technologies offer efficient means of fostering advancements in plant variety development. Processing large amounts of multidimensional breeding data over generations is complex, even though breeding information technologies provide an accessible and scientifically sound approach. Therefore, decision support tools help breeders to extract relevant and valuable information by introducing the golden seed breeding cloud platform. This platform is a cutting-edge AI driven agriculture system designed to maximize seed breeding for increased sustainability, resilience and yield. This brought a paradigm shift in farming by developing AI-powered solutions. To accomplish data integration and feature identification for stress phenotyping, it is necessary to take advantage of machine learning algorithms to extract patterns and features from the massive repository of data. Currently, plant breeding is propelling a revolution driven by state-of-the-art amenities for crop phenotyping and genome sequencing. Advanced phenotyping and genotyping when coupled with machine learning and cognitive sciences, are improving the accuracy of identifying the underlying genetic causes of attributes. Another advancement i.e., Next-generation AI using big data envisages how it can deal with the challenges by interfacing with the multi-omics big data to accelerate plant breeding, particularly for climate-resilient agriculture. Successful implementation of the proposed model based on big data characteristics will facilitate the evolution of breeding from “art” to “science” and eventually to “intelligence” in the era of Artificial Intelligence.

The growth rate of global grain production can no longer meet the demands arising from population expansion. Meanwhile, climate change, cultivated land degradation and environmental stress are further exacerbating the vulnerability of the food production system. Ensuring food security is fundamental to maintain the stability and sustainable development of human society. Crop breeding is the pivotal technical means to achieve this goal. Developing new crop varieties with high yield, superior quality, and multiple resistances have become the core pathway to safeguard agricultural sustainable development and global food supply. However, traditional breeding techniques suffer from long cycles, low accuracy and limited efficiency, which cannot satisfy the development requirements of modern seed industry. The rapid iteration of artificial intelligence (AI) technology has injected new intelligent momentum into the innovation of crop breeding, driving the transformation of breeding technology from traditional experience-based breeding to precise, intelligent and efficient breeding. In this review, we summarized the development of crop breeding, and focused on the innovative breakthroughs of genomic selection, precision genome editing, protein design and high-throughput phenotyping driven by AI. We also further elaborated the intelligent driving effects exerted by these technologies on key breeding links involving germplasm mining, gene function analysis, directional trait improvement, intelligent phenotypic assessment and intelligent factory breeding. Finally, we discussed the challenges and future developmental prospects of AI deployment in crop breeding, aiming to provide a reference for the innovation and industrial application of intelligent breeding technologies.

Global food security is confronted with mounting pressures arising from continuous population growth, intensifying climate change, and increasing agro-ecological vulnerability. Coarse grain crops, characterized by strong tolerance to drought and poor soil conditions, serve as strategic reserve crops with high nutritional value in the development of diversified food systems. Nevertheless, insufficient research investment and underdeveloped breeding infrastructure have led to the long-term marginalization of many coarse-grain crop resources. Smart breeding, through the integration of biotechnology and artificial intelligence (AI), opens new pathways for breaking through the bottlenecks in coarse grain crops breeding. This review systematically outlined the evolution of smart breeding technologies and discussed the applications of biological big data and AI in the digitization of coarse-grain crop germplasm resources, high-throughput phenomics, genotype-phenotype association analysis, and intelligent decision-making systems. We synthesized representative smart breeding cases in oats, foxtail millet, buckwheat, quinoa, and other coarse grain crops, explored pathways for the industrialization of coarse-grain breeding, and identified persistent challenges—specifically, the shortfall in fundamental research, institutional frictions in the breeding innovation system, and structural constraints in human resources and funding mechanisms and proposed corresponding countermeasures. Finally, we provided an outlook on frontier directions including de novo domestication, genomic design breeding, and digital twins breeding.

Genomic selection is a revolutionary approach in breeding, exploiting genetic markers to forecast breeding values and hence accelerating the pace of traits associated with resilience, like drought tolerance, heat resistance, and pest resistance. This study addresses these challenges through ML algorithms such as random forests, support vector machines, and neural networks thereby enhancing prediction accuracy while handling complicated genomic as well as environmental datasets. Relevant ML algorithms for genomic selection are considered in this discussion, as well as strategies for data processing, feature selection, and environmental factors, including climate conditions and soil parameters. These are brought together to form predictive models that indeed cater to genotype-by-environment interactions vital for crop performance evaluation over different environmental conditions. A proposed framework integrates genomic selection with machine learning, benefiting both disciplines by developing a data-driven methodology for yield prediction in corn. The critical machine learning models to be used include multi-layer perceptron and ensemble models. A case study shows the practical applicability of the GS-ML framework, describing the dataset prepared, model testing and validation procedures, and yield resilience prediction results. The conclusion of the study states that GS and ML combined hold great promise in supporting sustainable agriculture and climate resilience. It requires further research, infrastructure development, and policy support to scale this approach across different crops and diverse climate scenarios. The combined use of genomic and ML approaches is profoundly innovative in predictive breeding and will help develop resilient agricultural systems critical for global food security under a changing climate.

Genomic selection has been considered as a novel methodology beyond traditional marker assisted selection methods (MAS). GS can be considered as a variant of MAS selecting favourable individuals largely based on estimated breeding values derived genomically. It involves genotyping markers and phenotyping individuals in reference population then predicting phenotypes of candidates for selection using statistical machine learning models. New candidate individuals get predictions performed on them post trained model output if genotypic information happens to be available somehow. Selection of training population proves highly crucial for testing purposes and ultimately determines accuracy in genomic selection processes. Genomic selection models frequently utilize involve stepwise regression, ridge regression, genomic best linear unbiased prediction, Ridge regression best linear unbiased Prediction, Bayes A, Bayes B, Bayes care Bayesian model and least absolute shrinkage selection operator. This review aims to present an overview of genomic selection as an advanced breeding strategy that integrates genome wide markers and statistical model to accelerate genetic improvement in plants. It highlights the principles, methods and applications of genomic selection in enhancing crop traits and breeding efficiency.

… of genomic selection (GS) and genome editing (GE) for the improvement of complex traits in crop plants. The genetic architecture of polygenic traits, the principles underlying genomic …

Crop improvement is a long-term, expensive institutional endeavor. Genomic selection (GS), which uses single nucleotide polymorphism (SNP) information to estimate genomic breeding values, has proven efficient to increasing genetic gain by accelerating the breeding process in animal breeding programs. As for crop improvement, with few exceptions, GS applicability remains in the evaluation of algorithm performance. In this study, we examined factors related to GS applicability in line development stage for grain yield using a hard red winter wheat (Triticum aestivum L.) doubled-haploid population. The performance of GS was evaluated in two consecutive years to predict grain yield. In general, the semi-parametric reproducing kernel Hilbert space prediction algorithm outperformed parametric genomic best linear unbiased prediction. For both parametric and semi-parametric algorithms, an upward bias in predictability was apparent in within-year cross-validation, suggesting the prerequisite of cross-year validation for a more reliable prediction. Adjusting the training population’s phenotype for genotype by environment effect had a positive impact on GS model’s predictive ability. Possibly due to marker redundancy, a selected subset of SNPs at an absolute pairwise correlation coefficient threshold value of 0.4 produced comparable results and reduced the computational burden of considering the full SNP set. Finally, in the context of an ongoing breeding and selection effort, the present study has provided a measure of confidence based on the deviation of line selection from GS results, supporting the implementation of GS in wheat variety development.

To increase genetic gain for tolerance to drought, we aimed to identify environmentally stable QTL in per se and testcross combination under well-watered (WW) and drought stressed (DS) conditions and evaluate the possible deployment of QTL using marker assisted and/or genomic selection (QTL/GS-MAS). A total of 169 doubled haploid lines derived from the cross between CML495 and LPSC7F64 and 190 testcrosses (tester CML494) were evaluated in a total of 11 treatment-by-population combinations under WW and DS conditions. In response to DS, grain yield (GY) and plant height (PHT) were reduced while time to anthesis and the anthesis silking interval (ASI) increased for both lines and hybrids. Forty-eight QTL were detected for a total of nine traits. The allele derived from CML495 generally increased trait values for anthesis, ASI, PHT, the normalized difference vegetative index (NDVI) and the green leaf area duration (GLAD; a composite trait of NDVI, PHT and senescence) while it reduced trait values for leaf rolling and senescence. The LOD scores for all detected QTL ranged from 2.0 to 7.2 explaining 4.4 to 19.4% of the observed phenotypic variance with R2 ranging from 0 (GY, DS, lines) to 37.3% (PHT, WW, lines). Prediction accuracy of the model used for genomic selection was generally higher than phenotypic variance explained by the sum of QTL for individual traits indicative of the polygenic control of traits evaluated here. We therefore propose to use QTL-MAS in forward breeding to enrich the allelic frequency for a few desired traits with strong additive QTL in early selection cycles while GS-MAS could be used in more mature breeding programs to additionally capture alleles with smaller additive effects.

Climate change and environmental stresses pose severe, multifaceted risks to global food security and environmental sustainability, and result in the loss of primary productivity and biodiversity that negatively impact the environmental and socio-economic conditions of affected regions. Among other approaches, priming constitutes an easy and relatively cheap strategy due to its potential to enhance germination and stress resilience under changing environments. This review examines an emerging shift in crop improvement, in which environmental stress is no longer viewed solely as a constraint but also as a potential tool for enhancing plant resilience through stress priming and molecular memory. Various priming strategies applied through methods such as hydropriming, osmopriming, hormonal priming, chemical priming, thermopriming, biopriming, and nanopriming, effectively enhance germination performance and stress tolerance through activated defense pathways, osmolyte accumulation, and antioxidant system modulation. Advances in transcriptomics, metabolomics, and proteomics have revealed key markers of the primed state, including gene expression changes, metabolite accumulation, and epigenetic programming, which can provide tools for selection. These markers offer valuable opportunities for identifying and selecting genotypes with enhanced priming responsiveness. Integrating priming technologies with modern breeding strategies, particularly genomic selection, may therefore provide a powerful framework for improving stress adaptation in crops. By combining physiological priming with advanced genomic tools, this approach offers a practical and cost-effective route to accelerate the development of climate-resilient crop varieties and support sustainable agricultural production under increasingly variable environmental conditions.

Traditional breeding strategies for selecting superior genotypes depending on phenotypic traits have proven to be of limited success, as this direct selection is hindered by low heritability, genetic interactions such as epistasis, environmental-genotype interactions, and polygenic effects. With the advent of new genomic tools, breeders have paved a way for selecting superior breeds. Genomic selection (GS) has emerged as one of the most important approaches for predicting genotype performance. Here, we tested the breeding values of 240 maize subtropical lines phenotyped for drought at different environments using 29,619 cured SNPs. Prediction accuracies of seven genomic selection models (ridge regression, LASSO, elastic net, random forest, reproducing kernel Hilbert space, Bayes A and Bayes B) were tested for their agronomic traits. Though prediction accuracies of Bayes B, Bayes A and RKHS were comparable, Bayes B outperformed the other models by predicting highest Pearson correlation coefficient in all three environments. From Bayes B, a set of the top 1053 significant SNPs with higher marker effects was selected across all datasets to validate the genes and QTLs. Out of these 1053 SNPs, 77 SNPs associated with 10 drought-responsive transcription factors. These transcription factors were associated with different physiological and molecular functions (stomatal closure, root development, hormonal signaling and photosynthesis). Of several models, Bayes B has been shown to have the highest level of prediction accuracy for our data sets. Our experiments also highlighted several SNPs based on their performance and relative importance to drought tolerance. The result of our experiments is important for the selection of superior genotypes and candidate genes for breeding drought-tolerant maize hybrids.

… to assess the predictive ability of genomic selection (GS) for … of GS into breeding schemes of these crops. For alfalfa, the … interest, given the long selection cycle and the low narrow-…

Genomic selection applied for wheat quality in CIMMYT spring bread wheat breeding program. All wheat quality traits predicted and validated using forward genomic selection. Dough and loaf traits have moderately high predictive ability in CIMMYT breeding program. Genomic selection genetic gain 1.4 to 2.7 times higher than phenotypic selection.

A thorough verification of the ability of genomic selection (GS) to predict estimated breeding values for pea (Pisum sativum L.) grain yield is pending. Prediction for different environments (inter-environment prediction) has key importance when breeding for target environments featuring high genotype × environment interaction (GEI). The interest of GS would increase if it could display acceptable prediction accuracies in different environments also for germplasm that was not used in model training (inter-population prediction). Some 306 genotypes belonging to three connected RIL populations derived from paired crosses between elite cultivars were genotyped through genotyping-by-sequencing and phenotyped for grain yield, onset of flowering, lodging susceptibility, seed weight and winter plant survival in three autumn-sown environments of northern or central Italy. The large GEI for grain yield and its pattern (implying larger variation across years than sites mainly due to year-to-year variability for low winter temperatures) encouraged the breeding for wide adaptation. Wider within-population than between-population variation was observed for nearly all traits, supporting GS application to many lines of relatively few elite RIL populations. Bayesian Lasso without structure imputation and 1% maximum genotype missing rate (including 6058 polymorphic SNP markers) was selected for GS modelling after assessing different GS models and data configurations. On average, inter-environment predictive ability using intra-population predictions reached 0.30 for yield, 0.65 for onset of flowering, 0.64 for seed weight, and 0.28 for lodging susceptibility. Using inter-population instead of intra-population predictions reduced the inter-environment predictive ability to 0.19 for grain yield, 0.40 for onset of flowering, 0.28 for seed weight, and 0.22 for lodging susceptibility. A comparison of GS vs phenotypic selection (PS) based on predicted genetic gains per unit time for same selection costs suggested greater efficiency of GS for all traits under various selection scenarios. For yield, the advantage in predicted efficiency of GS over PS was at least 80% using intra-population predictions and 20% using inter-population predictions. A genome-wide association study confirmed the highly polygenic control of most traits. Genome-enabled predictions can increase the efficiency of pea line selection for wide adaptation to Italian environments relative to phenotypic selection.

Although long-term genetic gain has been achieved through increasing use of modern breeding methods and technologies, the rate of genetic gain needs to be accelerated to meet humanity's demand for agricultural products. In this regard, genomic selection (GS) has been considered most promising for genetic improvement of the complex traits controlled by many genes each with minor effects. Livestock scientists pioneered GS application largely due to livestock's significantly higher individual values and the greater reduction in generation interval that can be achieved in GS. Large-scale application of GS in plants can be achieved by refining field management to improve heritability estimation and prediction accuracy and developing optimum GS models with the consideration of genotype-by-environment interaction and non-additive effects, along with significant cost reduction. Moreover, it would be more effective to integrate GS with other breeding tools and platforms for accelerating the breeding process and thereby further enhancing genetic gain. In addition, establishing an open-source breeding network and developing transdisciplinary approaches would be essential in enhancing breeding efficiency for small- and medium-sized enterprises and agricultural research systems in developing countries. New strategies centered on GS for enhancing genetic gain need to be developed.

… Genomic selection (GS) will shorten the crop breeding cycle by identifying and tracking … The first section summarizes genomic selection and the contemporary phenotypic selection and …

… Toward integration of genomic selection with crop modelling: the development of an integrated approach to predicting rice heading dates. Theoretical and Applied Genetics 129(4): 805–…

… Thus, genomic selection (GS) is a promising method that utilizes genome-wide markers and can be applied to oilseed crops widely in the selection of complex traits in plant breeding …

Genomic selection (GS) has become an important tool for accelerating genetic gain in wheat breeding by enabling the prediction of target traits using genome-wide molecular markers. However, the large-scale implementation of GS in public breeding programs remains constrained by the cost of high-density genotyping platforms. Medium-density targeted genotyping approaches provide a cost-effective alternative while maintaining prediction accuracy. In this study, we evaluated the performance of a public wheat mid-density genotyping platform (Wheat DArTag 3.9K EIB 2.0) for GS by comparing it with a previously deployed higher-density genotyping-by-sequencing (GBS) platform. The analyses were conducted using five consecutive years of CIMMYT Elite Yield Trials comprising more than 5000 elite spring wheat lines evaluated across multiple irrigated, drought, and heat-stressed environments. Trait predictability was assessed for agronomic, phenological, and disease resistance traits using the genomic best linear unbiased prediction (GBLUP) model under several cross-validation scenarios, including within-year and across-year predictions. After quality filtering, the DArTag platform retained approximately 1600–1800 SNPs, whereas the GBS platform retained approximately 6500–9600 SNPs. Across traits and years, no consistent superiority of GBS over DArTag was observed, and correlations between genomic estimated breeding values (GEBVs) obtained from both platforms were high, indicating that both genotyping systems would lead to highly similar selection decisions. The results suggest that, in elite wheat germplasm characterized by long-range linkage disequilibrium and strong realized genomic relationships, medium-density targeted genotyping platforms can retain most of the predictability achieved by higher-density systems. Overall, the public Wheat DArTag 3.9K EIB 2.0 platform represents a scalable and cost-effective solution for implementing GS in operational wheat breeding programs.

… necessitate whole-genome prediction models and selection methodology as … whole-genome prediction models and genomic selection for disease resistance. In general, whole-genome …

… of genomic selection. It further discusses relevant prerequisites for the application of genomic selection, … this chapter), before concluding with some general notes on genomic selection. …

… Towards genomic selection for crop species GS has the potential to improve existing breeding schemes in plants and livestock. We postulate that improved breeding using new …

… disease resistance, especially focusing on the discovery of genetic markers for cattle disease resistance … for the research of animal disease resistance, promote the healthy development …

ABSTRACT Introduction: Antibiotic resistance (AR) is a worldwide concern and the description of AR have been discovered mainly because of their implications in human medicine. Since the recent burden of whole-genome sequencing of microorganisms, the number of new AR genes (ARGs) have dramatically increased over the last decade. Areas covered: In this review, we will describe the different methods that could be used to characterize new ARGs using classic or innovative methods. First, we will focus on the biochemical methods, then we will develop on molecular methods, next-generation sequencing and bioinformatics approaches. The use of various methods, including cloning, mutagenesis, transposon mutagenesis, functional genomics, whole genome sequencing, metagenomic and functional metagenomics will be reviewed here, outlining the advantages and drawbacks of each method. Bioinformatics softwares used for resistome analysis and protein modeling will be also described. Expert opinion: Biological experiments and bioinformatics analysis are complementary. Nowadays, the ARGs described only account for the tip of the iceberg of all existing resistance mechanisms. The multiplication of the ecosystems studied allows us to find a large reservoir of AR mechanisms. Furthermore, the adaptation ability of bacteria facing new antibiotics promises a constant discovery of new AR mechanisms.

Numerous environmental reservoirs contribute to the widespread antibiotic resistance problem in human pathogens. One environmental reservoir of particular importance is the intestinal bacteria of food-producing animals. In this review I examine recent discoveries of antibiotic resistance genes in agricultural animals. Two types of antibiotic resistance gene discoveries will be discussed: the use of classic microbiological and molecular techniques, such as culturing and PCR, to identify known genes not previously reported in animals; and the application of high-throughput technologies, such as metagenomics, to identify novel genes and gene transfer mechanisms. These discoveries confirm that antibiotics should be limited to prudent uses.