《1 Engineering research fronts》

1 Engineering research fronts

《1.1 Trends in Top 10 engineering research fronts》

1.1 Trends in Top 10 engineering research fronts

The Top 10 engineering research fronts in the field of agriculture mainly include: ① research on biological mechanisms and mechanisms related to crop breeding, such as “crop pan-genome” and “functional gene identification by multi-omics in animals”;② research on improving the quality, yield, and green production of animal and plant products, such as “mechanisms and methods for synergistic improvement of crop yield, quality, and efficiency”, “genetic basis and regulatory network for the quality formation of horticultural crops”, “intelligent identification mechanism and real-time monitoring technology for crop diseases and pests”, “diagnosis of forest diseases and pests based on deep learning”, and “straw modification and rapid decomposition technology”; ③ animal medicine and nutrition are still at the forefront of research in animal husbandry, such as “mechanisms of host inflammatory response regulation mediated by significant animal pathogens”, as well as “antibiotic-free nutritional regulation techniques for the intestinal health and growth of livestock and poultry”. The number of core papers in the engineering research fronts in agriculture ranges from 18 to 67, with an average of 47 papers, similar to previous years. The citations per paper range of articles is 28.94−225.40, with an average of approximately 84.54 times. The core papers are mainly published in 2018 and 2019, with the average publication year of core papers on “intelligent identification mechanism and real time monitoring technology for crop diseases and pests”, “straw modification and rapid decomposition technology”, “antibiotic-free nutritional regulation techniques for the intestinal health and growth of livestock and poultry”, “intelligent collaborative operation technology for multiple agricultural machinery”, and “diagnosis of forest diseases and pests based on deep learning” being 2019, compared to other selected fronts, it is closer to the current situation (Tables 1.1.1 and 1.1.2).

《Table 1.1.1》

Table 1.1.1 Top 10 engineering research fronts in agriculture

《Table 1.1.2》

Table 1.1.2 Annual number of core papers published for the Top 10 engineering research fronts in agriculture

(1) Crop pan-genome

A pan-genome is a collection of all DNA sequences of a species. In recent years, crop pan-genomics has become an essential field of the studies in crop genomics. It plays a key role in revealing the genetic variations, evolutionary histories and functional genes in crops. Currently, all the staple crops, such as maize, rice and wheat, have been assembled with multiple high-quality reference genomes. Comparing to the traditional single reference genome for a species, it has become obvious that multiple genome assemblies of different individuals or varieties will provide more adequate and accurate information in presenting the genetic diversity of the species, especially reveal the rich structure variations, such as tandem repeats, present/absent variations and chromosome translocations. Constructing pan-genome at genus level will provide valuable evidence for understanding on crop origin and evolution, decoding the domestication and de-domestication process, and revealing the evolution process of functional genes. Moreover, integrating the resourceful sequence information provides opportunity for resolving the core genes and dispensable genes at pan-genome level, promoting the mining and exploiting of beneficial gene recourses. In recent years, the achievement of super-high-quality genome assemblies, such as gap-less genome and telomere-to-telomere complete genome assemblies provide strong guidance in constructing crop pan-genome. The methodological method and analysis approaches based on pan-genomes are unceasingly improving, and the tasks such as the sequence mapping on pan-genome, functional annotation of gene family, and constructing graph-based pan-genome are calling for technical innovations. In the future, incorporating with advanced technologies, such as artificial intelligence and machine learning, and abundant phenotype data, the crop pan-genomics is expected to serve as a uniform and comprehensive coordinate system for facilitating gene mining and molecular design breeding, ultimately accelerate the crop basic research and development of improved varieties for ensuring national food security and sustainable agricultural development.

(2) Mechanisms and methods for synergistic improvement of crop yield, quality, and efficiency

The coordinated improvement of crop yield, quality and efficiency refers to the systematic revelation of the rules and mechanisms underlying the synergistic formation of high yield, superior quality, and high efficiency. This involves selecting suitable varieties and optimal planting times based on local conditions to make full use of local temperature and light resources, optimizing nutrient management and irrigation to achieve efficient utilization of fertilizer and water resources. As a result, it accomplishes the synergistic enhancement of crop yield, quality, and efficiency. At present, the yield per unit area of major crops such as rice, wheat, and corn in China has reached a relatively high level. However, with the rapid development of China’s economy and society as well as the continuous improvement in people’s quality of life, the modern high-quality development of agriculture poses new requirements for crop production. There is an urgent need to not only focus on increasing yield per unit area but also to effectively enhance quality and resource utilization efficiency. Therefore, it is crucial to analyze the mechanisms and explore the pathways for the coordinated improvement of crop yield, quality, and efficiency. The core scientific issues in this regard mainly include:

1) The interplay between crop growth, development, yield, quality and efficiency, the comprehensive effects of crop-environment- cultivation measures on yield and quality efficiency, and the mechanisms underlying the coordinated improvement of crop yield and quality efficiency.

2) The evaluation index system and comprehensive selection methods for crop varieties that promote coordinated improvement in yield, quality and efficiency.

3) The regulatory pathways and key technologies that can be widely applied for the coordinated improvement of crop yield, quality and efficiency. Breakthroughs in the relevant theories and technologies mentioned above will provide essential theoretical support and modern engineering solutions for China’s coordinated production of high yield, superior quality, and high efficiency crops.

(3) Genetic basis and regulatory network for the quality formation of horticultural crops

As the horticultural industry develops rapidly and the total output of horticultural products increases significantly, changes in product quality, health component content, and quality and safety issues of horticultural products have received close attention. The quality of horticultural products includes factors such as appearance, nutritional quality, and flavor quality, mainly including organ size, shape, uniformity, color, storage and transport ability, various nutrient contents (sugar, acid, starch, vitamins, mineral nutrients, flavonoids, etc.), various flavor substance contents (amino acids, aromatic substances, etc.), and taste (softness, hardness, and flavor perception). The quality traits are complex. In recent years, some research has been carried out in China on fruit shape, color, nutritional quality, flavor, and the formation and regulation mechanism of bitter substances in horticultural crops. In particular, the use of genomics, transcriptomics, metabolomics, and other methods has clarified the key regulatory genes for a few metabolic substances by exploring and analyzing the metabolic genes and their regulatory genes related to horticultural product quality. There is also an increasing understanding of the molecular mechanisms and regulatory networks of important agronomic traits. However, many metabolic substances that determine the quality of horticultural products are still unclear, and there is limited research on the molecular mechanisms and metabolic regulatory mechanisms. Therefore, in the future, different levels of gene expression cascade regulatory mechanisms and their regulatory networks, as well as the interactive regulation between plant hormone signal transduction and quality formation, the coupling regulation mechanism between quality metabolism and the environment, and other research should be carried out based on the existing foundation. It aims to explore the control network and signal transduction mechanism of horticultural product quality formation, and provide a scientific basis for improving the quality and efficiency of the horticultural production.

(4) Intelligent identification mechanism and real-time monitoring technology for crop diseases and pests

In the system for the integrated pest management, control decision is the core, prediction is the basis of decision-making, and monitoring is an indispensable basis for prediction and decision-making. Nowadays, timely, rapid, and accurate monitoring technology is very important in plant disease and pest control. With the integration and development of high and new technologies such as remote sensing, radar, image technology, quantitative molecular detection technology, sensor, Internet of Things technology, big data analysis and artificial intelligence in recent years, relatively obvious progress has been made in the construction of intelligent identification, real-time monitoring and early warning technology of crop diseases and pests, which has greatly improved the timeliness and accuracy of disease and insect monitoring. The intelligent identification technology of crop diseases and pests is mainly composed of several or several modules, such as capture and lure equipment, detection or high-definition shooting equipment, real-time transmission platform, remote monitoring platform, intelligent identification and counting model or algorithm, and network client of computer or mobile phone, the monitoring of pests and diseases has been transformed from the traditional manual investigation to the integrated remote automatic real-time monitoring, identification and diagnosis of real-time collection, transmission, identification, analysis and early warning of pest and disease information. Although the accuracy or application scope of most identification and monitoring models is inevitably limited, with the further development of computer technology such as artificial intelligence and deep learning, the intelligent identification and monitoring technology of crop pests and diseases integrated with traditional methods and emerging technologies will become the main development direction of disease and pest monitoring and early warning in the future.

(5) Straw modification and rapid decomposition technology

Straw return is important for maintaining soil fertility, improving nutrient efficiency, and slowing soil acidification. However, due to the high carbon to nitrogen ratio and the complex structure of straw components, direct straw return is not easy to decompose quickly, which may cause nitrogen deficiency, aggravate the diseases, and influence the growth of the next crop. Ectopic decomposition of straw also takes at least 2 months. Therefore, how to promote the rapid decomposition of straws is one of the urgent treatment problems in current agriculture. The process of straw decomposition to organic matter was influenced by the nature of the straw itself, soil chemical and soil microbial properties. Straw modification refers to the exogenous addition of nitrogen fertilizers, nanomaterials, chemical modifiers, enzyme preparations or microbial inoculants, and so on, which can regulate the physical, chemical, biological processes during the straw decomposition, and thus improve the decomposition efficiency of straw and product efficacy. With the rapid development of techniques such as microbiology and organic substance characterization, it can be achieved to increase the speed of in-situ or ectopic straw decomposition, improve the performance of organic products and nutrient capacity based on techniques of highly efficient microbial inoculant and straw modification, and finally establish regional precision technology for rapid decomposition of straw under different climatic environments. This is of great significance for realizing the sustainable and low-carbon development of agriculture, and accelerating the agricultural recycling economy.

(6) Mechanisms of host inflammatory response regulation mediated by significant animal pathogens

As China is the world‘s largest consumer of livestock and poultry and animal-derived food, the prevention and control of significant animal diseases are crucial for the healthy development of animal husbandry and public health security. The invasion of animal pathogens induces the host immune response, and the regulation of host inflammatory response is closely related to the occurrence and development of diseases. Inflammatory response is a self-defense mechanism of the host against stimuli, playing an important role in pathogen clearance, enhancing disease resistance, and preventing the spread of pathogens. Specifically, upon invading the host, pathogens are recognized by host pattern recognition receptors, initiating a series of intracellular signal transduction, recruiting a large number of chemokines, cytokines, and activated immune cells to gather at the infected site to clear the pathogens, causing local inflammatory reactions. Inflammatory response is mediated by pro-inflammatory cytokines, and some important pathogens can induce a large release of pro-inflammatory cytokines to cause “cytokine storm”, leading to strong local inflammation of tissues and assisting pathogen proliferation; meanwhile, pathogens can cause macrophage lysis by activating the inflammasome, resulting in the escape and spread of pathogens. In addition, pathogens can inhibit the expression and release of pro-inflammatory factor by regulating various signaling pathways in host cells, thereby avoiding clearance by the host and establishing persistent infection. Elaborating the key mechanisms by which pathogens regulate host inflammatory response is crucial for understanding the pathogenicity of pathogens and seeking new treatment of infectious diseases. The relevant research results will also provide important guidance and theoretical basis for vaccine development and the prevention and control of significant animal diseases.

(7) Antibiotic-free nutritional regulation techniques for the intestinal health and growth of livestock and poultry

The problems of antibiotic resistance, ecological environment pollution and food safety caused by the abuse of antibiotics have seriously affected the healthy development of animal husbandry in China. In 2020, China issued a “prohibition order” to completely prohibit the addition of growth-promoting antibiotics in feed, which brings great challenges to intestinal health and growth and development of livestock and poultry. Based on this, the development and application of green and safe non- resistant nutrition control technology to maintain the intestinal health of livestock and poultry and promote their growth and development is an inevitable trend of efficient and sustainable development of animal husbandry in China. The non- resistant nutrition regulation technology is based on the balance of feed “nutrient structure”, integrating integrated nutrients to the comprehensive nutritional regulation of animal immunity, intestinal health and pathogenic factors, so as to promote the intestinal development of animals under the condition of non-resistance, improve the body’s resistance to disease, and ensure animal health and efficient production. The key of non-resistant nutrition technology system lies in the balance and optimization of feed nutrient structure, feed processing modulation and precise feeding, regulating intestinal microecological balance from the aspects of nutrient source selection, formula design and optimization, additive combination screening, feed processing modulation, feeding methods and fecal bacteria transplantation, so as to promote intestinal health and efficient growth of livestock and poultry. It is difficult and hot to further strengthen the basic theoretical research of non-resistant nutrition and promote the application of non-resistant nutrition regulation technology system. The relevant research results will certainly provide theoretical guidance and technical support for the intestinal health and efficient growth of livestock and poultry under non-resistant conditions.

(8) Functional gene identification by multi-omics in animals

Functional gene identification by multi-omics in animals is a process that involves the comprehensive analysis of various genetic, transcriptomic, and proteomic data, along with related bioinformatics methods, to identify and study genes in the animal genome that have important functions and biological significance. The process of functional gene identification in animal multi-omics typically includes the following steps: data collection, data preprocessing, bioinformatics analysis (which involves the application of bioinformatics tools and algorithms to analyze the preprocessed data, including genome assembly, gene localization, gene function annotation, transcriptome assembly, expression analysis, protein identification and quantification), functional gene mining (which involves the identification of genes with important functions and biological significance by combining the analysis results with existing biological knowledge, and may include enrichment analysis, differential expression analysis, protein- protein interaction network analysis, etc.), functional annotation and interpretation, and biological validation (which includes experimental validation or further functional studies to confirm the important roles of functional genes in biological processes). This technique aims to decipher the molecular genetic basis of important traits such as meat quality and quantity, milk production and quality, wool production and quality, egg production, as well as growth, development, reproduction, disease resistance, heat stress resistance, and low-oxygen tolerance in superior animal breeds. It provides a basis for the theory and technological innovation of genome-guided breeding, greatly improves the accuracy of selecting and creating superior traits such as high yield, good quality, disease resistance, and stress tolerance in agricultural animals, and accelerates the process of targeted breeding of improved varieties.

(9) Intelligent collaborative operation technology for multiple agricultural machinery
According to statistics, the global population is projected to reach 9.7 billion by 2050, with a growth in food demand exceeding 70%. Labor shortages have become a pressing issue in agricultural production, necessitating the advancement of agricultural modernization and industrial upgrading. Unmanned intelligent agricultural machinery, through automation technology, can operate autonomously to reduce reliance on manual labor and achieve high-efficiency agricultural operations, thereby enhancing production efficiency. However, as agricultural modernization progresses, the expansion of farmland continues, and achieving intelligent and unmanned operation for a single agricultural machine may not suffice to meet the demands of large-scale, efficient production. The realization of intelligent collaborative operations among multiple agricultural machines has become a research focal point. This includes communication and networking technologies facilitating real-time data transmission and information exchange between machines, technologies for perceiving and identifying real-time agricultural environmental conditions and crop growth statuses, autonomous planning and coordinated control technologies for multiple agricultural machines, task coordination and operation scheduling technologies among these machines, safety and security technologies for both single and multiple agricultural machine collaborative operations—covering aspects such as safe maneuvering, data security, system reliability, and fault diagnosis, among others. Breaking through the core key technologies related to these aspects will enable task specialization and collaboration in agricultural production, leading to improved operational efficiency and supporting the development of smart agricultural applications.

(10) Diagnosis of forest diseases and pests based on deep learning

Deep learning-based approaches for forest pest and disease diagnostics represent an innovative strategy for detecting and monitoring these threats in woodland environments. Deep learning, a subset of machine learning techniques, centers on constructing and training multi-layered neural networks to autonomously glean features and patterns from extensive datasets. Such methods excel in accurately recognizing pests and diseases and scrutinizing their prevalence trends by synergizing data. This approach offers several advantages over conventional methods of identification, including alleviating the repetitive tasks undertaken by professionals, conserving labor, providing timely insights, and delivering high levels of precision and efficiency. Current research in the realm of forest pest and disease diagnosis utilizing deep learning encompasses various dimensions. These encompass image recognition for identifying forest pests and diseases, tracking pest populations and their dynamics via the detection of light- or pheromone-induced responses. Furthermore, this approach extends to monitoring alterations in the prevalence of forest pests and diseases and predicting their trajectory using high-resolution imagery from satellites and unmanned aerial vehicles. These images encompass visible, multispectral, and hyperspectral data. As artificial intelligence technology continues to advance, constructing a comprehensive model for forest pests and diseases, delving into data fusion and multimodal information, exploring model design, compression, and automated architecture searches are imperative. These measures collectively enhance the efficiency and adaptability of the model. Predicting pest and disease occurrences based on regional monitoring networks is pivotal in adapting to contemporary environmental shifts and combating the invasion of pests and diseases. This is particularly vital for ensuring the sustainable development of forestry under prevailing mega-trends.

《1.2 Interpretations for three key engineering research fronts》

1.2 Interpretations for three key engineering research fronts

1.2.1 Genetic basis and regulatory network for the quality formation of horticultural crops

With the rapid development of the horticulture industry and the substantial increase in total production of horticultural crops, there has been a growing concern for changes in product quality characteristics, content of health-promoting compounds, and the safety of horticultural products. Horticultural product quality includes aspects such as appearance, nutritional quality, and flavor quality, which mainly involve organ size, shape, uniformity, color, shelf-life, content of various nutrients (sugar, acid, starch, vitamins, mineral nutrients, flavonoids, etc.), content of various flavor compounds (amino acids, aromatic substances, etc.), and sensory attributes (texture, mouthfeel, etc.). The characteristics of product quality are complex. In recent years, China has conducted research on fruit shape, color, nutritional quality, flavor, and the formation and regulation mechanisms of bitter compounds in horticultural crops. Particularly, methods such as genomics, transcriptomics, and metabolomics have been applied to identify and analyze genes related to the metabolism of substances affecting horticultural product quality. This research has shed light on the key regulatory genes for a small number of metabolites and has led to a better understanding of the molecular mechanisms and regulatory networks underlying important agronomic traits. However, the molecular mechanisms and metabolic regulation of most metabolites that determine horticultural product quality remain unclear, with limited research in this area.

With the rapid development of sequencing technology and the quick decrease in costs, the genomes of many horticultural plants have been sequenced, or updated to more high-quality genomes. Chinese scientists have made outstanding contributions in this field. According to incomplete statistics, genome sequencing has been completed for more than 50 horticultural crops, providing a foundation for breeding new and superior crop varieties through molecular design methods.

Currently, over 60 quantitative trait loci controlling sugar content have been identified. Additionally, proteins indirectly involved in sugar metabolism, such as sucrose transporters, starch synthesis enzymes, and vacuole processing enzymes, have been analyzed. The team led by YE Zhibiao successfully identified major-effect locus TFM6 (tomato fruit malate 6), an Al-activated malate transporter ALMT9, that regulates malate accumulation in tomatoes through association analysis (mGWAS). They confirmed that WRKY42 negatively regulates ALMT9 expression by binding to the W-box on the ALMT9 promoter, thereby inhibiting malate accumulation in tomato fruits. It has been reported that 237 loci in tomatoes are associated with enzyme-catalyzed reactions in the ascorbic acid metabolic pathway, and the metabolic network of genes and corresponding enzymes related to the ascorbic acid metabolic pathway has been preliminarily established.

Many genes involved in carotenoid biosynthesis pathways have been isolated, cloned, and functionally validated. Secondary metabolites such as flavonoids, alkaloids, phenylpropanoids, and aromatic compounds are also important components in the formation of nutritional and flavor quality in horticultural crops. In this field, China’s Institute of Agricultural Sciences and the University of Florida in the USA have collaborated to identify 33 key flavor compounds influencing consumer preferences. They have also discovered 49 key loci regulating the accumulation of flavor compounds, revealing the material basis and genetic improvement path of tomato flavor. This provides feasible breeding solutions for cultivating delicious tomatoes.

Due to the complexity of quality traits, research on the genetic basis and regulatory networks underlying trait formation is still in its early stages. The future focus of research will be on the application of molecular biology techniques to establish and optimize genetic transformation systems for horticultural crops, to uncover key genes related to crop quality, to elucidate the regulatory role of these genes in quality, and to establish a more precise molecular regulatory network by identifying important nodal regulatory genes. Efforts will be made to break the unfavorable gene linkage with quality traits, achieve simultaneous improvements in horticultural crop quality and other desirable traits such as stress resistance, and breed high-quality horticultural crop varieties. The mechanisms by which environmental factors and cultivation techniques regulate quality will be explored, and key technologies for quality horticultural crop cultivation in controlled environments will be established.

In the front of “genetic basis and regulatory network for the quality formation of horticultural crops”, the top three countries in terms of the publication of core papers (Table 1.2.1) are China (63.79%), the USA (27.59%), and Italy (12.07%). The citations per paper in this front is distributed from 46.25 to 93.67, of which the citations per paper of Israel and the USA exceeds 70.00. In terms of the distribution of research institutions (Table 1.2.2), Chinese Academy of Sciences (CAS), Anhui Agricultural University (AAU), and Chinese Academy of Agricultural Sciences (CAAS) have produced many core papers and cited many times. Cooperation among countries is more common, and cooperation among China and the USA is relatively closer (Figure 1.2.1). The output is mainly in the cooperation network among institutions (Figure 1.2.2), and there are certain cooperation relations among these institutions. The main output countries of the citing papers are China and the USA, with China accounting for 49.05% and the USA accounting for 14.35% (Table 1.2.3). In terms of the main output institutions of the citing papers (Table 1.2.4), CAS, CAAS, and Nanjing Agricultural University (NAU) ranked among the top three in the number of citing papers. Figure 1.2.3 shows the roadmap of the engineering research front of “genetic basis and regulatory network for the quality formation of horticultural crops”.

《Table 1.2.1》

Table 1.2.1 Countries with the greatest output of core papers on “genetic basis and regulatory network for the quality formation of horticultural crops”

《Table 1.2.2》

Table 1.2.2 Institutions with the greatest output of core papers on “genetic basis and regulatory network for the quality formation of horticultural crops”

《Figure 1.2.1》

Figure 1.2.1 Collaboration network among major countries in the engineering research front of “genetic basis and regulatory network for the quality formation of horticultural crops”

《Figure 1.2.2》

Figure 1.2.2 Collaboration network among major institutions in the engineering research front of “genetic basis and regulatory network for the quality formation of horticultural crops”

《Table 1.2.3》

Table 1.2.3 Countries with the greatest output of citing papers on “genetic basis and regulatory network for the quality formation of horticultural crops”

《Table 1.2.4》

Table 1.2.4 Institutions with the greatest output of citing papers on “genetic basis and regulatory network for the quality formation of horticultural crops”

《Figure 1.2.3》

Figure 1.2.3 Roadmap of the engineering research front of “genetic basis and regulatory network for the quality formation of horticultural crops”

1.2.2 Functional gene identification by multi-omics in animals

The healthy development of the livestock and poultry industry is a strategic necessity to ensure the safety of our country’s grain and livestock products. As a key component of the breeding industry, the seed industry will further contribute to the development of animal husbandry in the future. The molecular mechanisms underlying important economic traits in livestock and poultry are extremely complex. Currently, the incomplete annotation of functional genomic maps and the unclear understanding of key genes and regulatory mechanisms for excellent traits severely restrict the efficient utilization of germplasm resources, genetic improvement, and the creation of new breeds in livestock and poultry.

With the continuous development of modern biotechnology, livestock and poultry breeders have progressed from macroscopically describing the genetic laws of traits based on genetic power to analyzing the regulatory regions that influence quantitative traits using molecular markers, and now to directly deciphering the relationship between DNA sequence variations and complex traits. These known causative mutations of important traits can be directly applied to breeding practices, laying a solid foundation for the cultivation of superior breeds. Functional gene mining in animal multi- omics is crucial for deciphering the molecular genetic basis of important economic traits in livestock and poultry, as well as for genome sequencing and functional genomics research. It is also an important prerequisite for the development of superior animal breeds and has been a focus of research in the field of animal genetic breeding both domestically and internationally for a long time.

Since 2009, whole-genome selection technology based on genome-wide SNP chips has rapidly developed and been widely used in dairy cattle, beef cattle, pigs, and broilers breeding in developed countries in Europe and America. In 2014, Australian scientists initiated the international 1 000 Bull Genomes Project. In 2015, the European Union and the USA established the “International Animal Genomic Functional Annotation Project” to predict the important traits of livestock and poultry germplasm resources. Chinese scientists have led or participated in genome projects of different livestock and poultry species such as chickens, ducks, pigs, geese, yaks, and sheep. They have also jointly launched scientific programs like the Ten Thousand Bird Genomes Project and the Macro-genome Project with different countries, laying a foundation for elucidating the genetic mechanisms of important economic traits. In addition, domestic and foreign researchers have identified a group of differentially expressed genes, differentially expressed proteins, and differentially methylated genes related to important economic traits such as growth, meat quality, reproduction, and milk production. They have explored the regulatory relationships between them and made important discoveries in molecular mechanisms such as pig carcass traits, milk composition synthesis and mastitis resistance in dairy cows, and the development of Cashmere goat hair follicles. They have obtained whole-genome copy number variation maps for cattle, pigs, beef cattle, chickens, and ducks, and discovered multiple candidate genes related to milk production traits, body size traits, carcass traits, immune traits, and meat quality traits. The initiation of domestic and international animal genomic functional annotation projects (FAANG, FarmGTEx, etc.) has led to the elucidation of the molecular regulatory mechanisms of agricultural animal phenotypic traits in the era of genomics. Meanwhile, as functional gene mining in animal multi-omics continues to deepen in basic research applications, agricultural animal breeding is accelerating its transition from traditional conventional breeding to multi- omics intelligent design breeding.

Animal multi-omics functional gene mining technology mainly includes comprehensive analysis of big data in animal pan- genomics, genomics, epigenomics, spatiotemporal transcriptomics, proteomics, metabolomics, and phenomics. It integrates methods from genetics, genomics, bioinformatics, molecular biology, biochemistry, cell biology, and animal breeding to reveal panoramic multi-dimensional omics features of animals. It aims to identify genetic loci and molecular markers (QTL) related to traits such as high productivity, high quality, stress resistance, and disease resistance in animals. It explores key genetic variations, functional genes, regulatory sequences, and regulatory networks, revealing the interactions between genes, phenotypes, and the environment. It provides molecular genetic selection markers and manipulation targets for animal breeding. This technology aims to decipher the molecular genetic basis of important traits such as meat quantity and quality, milk yield and dairy quality, fur yield and quality, egg production, as well as growth, development, reproduction, disease resistance, heat stress tolerance, and other important traits in excellent animal germplasm. It provides a basis for genome- guided precision breeding theory and technological innovation, greatly improving the accuracy of selecting and creating high- yielding, high-quality, disease-resistant, and stress-tolerant traits in agricultural animals, and accelerating the targeted breeding process of superior breeds.

In terms of core papers published in the front of “functional gene identification by multi-omics in animals”, the top three countries are the USA (38.00%), China (38.00%), and the UK (10.00%) (Table 1.2.5). The citations per paper of this front ranges from 45.00 to 82.00, with the UK, Australia, and the Netherlands are all more than 65.00. In terms of the main output institutions, University of California, Davis, University of Toulouse, and Northwest A & F University produce core papers with higher citations per paper (Table 1.2.6). As shown in Figure 1.2.4, research cooperation between countries is relatively common, with the cooperation between the UK, the USA, and China being relatively closer. As shown in Figure 1.2.5, there is a certain degree of cooperation among main institutions. The top two countries in the number of citing papers are China and the USA, with China accounting for 30.38% and the USA accounting for 24.14%, and the mean year is relatively late, demonstrating strong research and development momentum (Table 1.2.7). In terms of the main institutions of citing papers (Table 1.2.8), CAS, CAAS, and China Agricultural University rank top three, with the number of citing papers of CAS ranking first.

《Table 1.2.5》

Table 1.2.5 Countries with the greatest output of core papers on “functional gene identification by multi-omics in animals”

《Table 1.2.6》

Table 1.2.6 Institutions with the greatest output of core papers on “functional gene identification by multi-omics in animals ”

《Figure 1.2.4》

Figure 1.2.4 Collaboration network among major countries in the engineering research front of “functional gene identification by multi-omics in animals”

《Figure 1.2.5》

Figure 1.2.5 Collaboration network among major institutions in the engineering research front of “functional gene identification by multi-omics in animals”

《Table 1.2.7》

Table 1.2.7 Countries with the greatest output of citing papers on “functional gene identification by multi-omics in animals”

《Table 1.2.8》

Table 1.2.8 Institutions with the greatest output of citing papers on “functional gene identification by multi-omics in animals”

It is worth noting that complex economic traits in animals are governed by a large number of genetic components, which are exhibited through intricate developmental regulatory networks and gradually shaped by natural or artificial selection. The contribution of a single gene to a trait is limited, and its effect can vary significantly under different environmental conditions. The regulatory networks that involve coupling and antagonism between different traits, as well as their underlying core regulatory units, have not been fully deciphered. The phenomena of multiple effects caused by genetic linkage and one effect caused by multiple factors have become key bottlenecks in animal multi-omics functional gene mining. These bottlenecks also limit molecular design breeding in animals, especially the genetic progress in multi-trait coordinated improvement. Therefore, future research directions should focus on two aspects: enriching the sources of multi-omics data and expanding the depth of multi- omics integrated analysis. ① Traditional functional genomics studies that focus on static genetic associations for single traits (such as SNP and GWAS), or functional studies of individual genes or signaling pathways during development, have difficulty accurately depicting the systemic evolutionary framework from genotype to phenotype. To address the unique multi-factor regulatory models underlying different economic traits, emerging integrated analysis technologies such as three-dimensional genomics (Hi-C), protein interaction networks, multi-dimensional epigenomics (ATAC-Seq, ChIP-seq), single-cell sequencing, spatial transcriptomics, gene editing techniques, and optimized GWAS algorithms can be fully utilized. These technologies can help unravel the regulatory networks formed by the combination and interaction of different alleles at different loci, which contribute to the formation of complex trait phenotypes. Key functional elements that can be used in breeding practices can be discovered. ② Molecular genetic studies of important traits in agricultural animals require the support of large-scale experimental populations and data, as well as continuous, systematic, and in-depth monitoring and research on key phenotypes. Therefore, the extensive research on multi-omics intelligent design breeding inevitably leads to an exponential growth of multi-omics information in animals. Interpreting massive and heterogeneous multi-omics data and integrating multi-omics information from different studies to analyze the relationship between genetic variations and important economic traits present significant challenges. Overcoming the bottleneck of single-omics analysis limited to “correlation” between markers and phenotypes and breaking through the difficulties in revealing “causality” can be achieved through reference to animal multi-omics databases and dynamic integrated analysis methods. From an integrative and systems biology perspective, the genetic basis of complex traits in animals can be

《Figure 1.2.6》

Figure 1.2.6 Roadmap of the engineering research front of “functional gene identification by multi-omics in animals”

elucidated. Figure 1.2.6 shows the roadmap of the engineering research front of “functional gene identification by multi-omics in animals”.

1.2.3 Diagnosis of forest diseases and pests based on deep learning

Forests, as intricate non-rigid organic entities within nature, play a pivotal role in ecosystems by shaping ecological niches, facilitating energy flow, and orchestrating material cycling. Furthermore, they stand as vital carbon sinks, harboring one of the largest repositories of organic carbon within terrestrial ecosystems. A report from the United Nations’ Food and Agriculture Organization (FAO) indicated that in 2020, global forest coverage spanned 4.045 billion hectares, constituting 31% of the Earth’s land area. Pairing this with insights from China’s comprehensive forest resources assessments and statistical data, the calculation of average forest benefits reveals that China’s per-unit forest area annually retains 192.34 tons of water, absorbs 12.1 kilograms of airborne pollutants, and releases 315 kilograms of oxygen. However, challenges persist, particularly concerning the health of forest ecosystems. In 2022, China encountered substantial threats from forestry pests, with 178 million mu (approximately 11.87 million hectares) of forest land imperiled by major pests. Pine wood nematode disease affected 22 million mu (approximately 1.47 million hectares) of forest land, resulting in 10 million diseased and dead pine trees. Additional threats, such as the white moth, imperiled another 28 million mu (approximately 1.87 million hectares). Consequently, research into diagnosing forest tree diseases and pests holds immense significance for upholding ecosystem health, enhancing forestry productivity, and safeguarding biodiversity.

Traditional methods of pest and disease identification in forests rely on manual techniques and conventional image processing methods. These approaches often suffer from limitations related to human judgment, constrained feature extraction capabilities, and compromised accuracy and efficiency. A scarcity of skilled technicians further exacerbates these issues, leading to substantial economic losses caused by delayed prevention and control efforts or excessive intervention that impacts ecological security and biodiversity.

Deep learning-based methods, on the other hand, capitalize on constructing intricate neural networks to autonomously grasp complex features and patterns from extensive datasets, resulting in more precise and rapid pest and disease diagnoses. These methods encompass a spectrum of algorithms, including multi-task learning for versatile diagnosis, transfer learning to minimize data requirements and training time, target detection algorithms to pinpoint pest and disease locations, and semantic segmentation to precisely delineate infected and healthy areas. Ongoing research within the domain of deep learning-based forest pest and disease diagnostics centers on several key facets. These include bolstering model efficiency and generalization, enabling real-time monitoring and feedback, enhancing automation and intelligence, leveraging data fusion and multimodal information, and harnessing the potential of AI’s large-scale models. This approach chiefly concentrates on areas such as forest pest and disease image recognition, population dynamics monitoring through techniques like lamp lure or pheromone-induced pest counting, trend monitoring in forest pest and disease areas using satellite imagery and unmanned aerial vehicle data (encompassing visible, multispectral, and hyperspectral information), and the prediction of pest and disease occurrence trends.

In the front of “diagnosis of forest diseases and pests based on deep learning”, the three leading countries in terms of the number of core papers published (Table 1.2.9) are the USA, China, and India. The citations per paper of this front ranges from 45.00 to 123.00, with the exceptions of Egypt and India, which exhibit a citation frequency surpassing 50.00. In terms of research institution distribution (Table 1.2.10), institutions such as COMSATS University Islamabad, Prince Sultan University, University of Washington, University of Hawaii, and University of Hawaii at Manoa have demonstrated a high output of core papers along with a high citations per paper. The primary countries contributing to the citing papers include China, India, and the USA (Table 1.2.11). Regarding the main institutions of citing papers (Table 1.2.12), CAS, COMSATS University Islamabad, and Deakin University hold prominent positions in terms of the number of citing papers. Collaboration between countries is extensive and interconnected (Figure 1.2.7), with relatively close collaborations observed between Egypt, India, and Saudi Arabia. As for the collaborative network among the main institutions (Figure 1.2.8), there exists a degree of cooperation both within institutions of the same country and across institutions in different countries. For instance, notable collaboration exists between Seoul National University and University of Waterloo, University of Hawaii and University of Hawaii at Manoa, and University of Washington, Prince Sultan University, and

《Table 1.2.9》

Table 1.2.9 Countries with the greatest output of core papers on “diagnosis of forest diseases and pests based on deep learning”

《Table 1.2.10》

Table 1.2.10 Institutions with the greatest output of core papers on “diagnosis of forest diseases and pests based on deep learning”

《Table 1.2.11》

Table 1.2.11  Countries with the greatest output of citing papers on “diagnosis of forest diseases and pests based on deep learning”

《Table 1.2.12》

Table 1.2.12 Institutions with the greatest output of citing papers on “diagnosis of forest diseases and pests based on deep learning”

《Figure 1.2.7》

Figure 1.2.7 Collaboration network among major countries in the engineering research front of “diagnosis of forest diseases and pests based on deep learning”

《Figure 1.2.8》

Figure 1.2.8 Collaboration network among major institutions in the engineering research front of “diagnosis of forest diseases and pests based on deep learning”

《Figure 1.2.9》

Figure 1.2.9 Roadmap of the engineering research front of “diagnosis of forest diseases and pests based on deep learning”

COMSATS University Islamabad. Figure 1.2.9 shows the roadmap of the engineering research front of “diagnosis of forest diseases and pests based on deep learning”.

《2 Engineering development fronts》

2 Engineering development fronts

《2.1 Trends in Top 11 engineering development fronts》

2.1 Trends in Top 11 engineering development fronts

The Top 11 engineering development fronts in the agricultural field mainly involves directions such as agricultural green development, smart agriculture, and agricultural engineering, and reflects interdisciplinary applications (Table 2.1.1). Among them, the engineering development fronts related to agricultural green development include “crop green super-high-yield cultivation technology”, “development and utilization of high-quality germplasm resources for horticultural crops”, “molecular design of green pesticides based on structural biology”, “synergistic technology for efficient conversion of organic matter and reduction of pollutants during composting”, and “bio-refinery of wood waste”. At the same time, the in-depth application of gene editing technology is a hot research topic for researchers, such as “key technologies for unmanned farms” and “ecological breeding technology of aquatic animals”. The disclosure of core patents related to various cutting-edge technologies from 2017 to 2022 is shown in Table 2.1.2. Among them, the citations per patent of “molecular design of green pesticides based on structural biology” is as high as 19.61, indicating that agricultural green development has received widespread attention from researchers in recent years. The core patents for “development and application of genome editors in crops” have the highest number of disclosures, with 4 658 published patents (up to 1 292 in 2022) and 63 027 citations, far higher than other development fronts. The core patents for “creation of novel and efficient animal vaccines” are the least, and there were no core patents in 2022.

(1) Development and application of genome editors in crops
Genome editing refers to the genetic manipulation technology of precise insertion, deletion or replacement of DNA. Through precise editing of functional genes or regulatory sequences in organisms, genome editing technology is able to repair target genes, strengthen beneficial genes, eliminate harmful genes, and ultimately achieve the purpose of improving the function of organisms. As the most subversive and leading cutting-edge technology of biological breeding, genome editing technology has become the priority in developed countries and multinational seed companies, and is the competition commanding height in the field of global agricultural biotechnology. Using genome editing technology, Chinese scientists have realized the genetic improvement

《Table 2.1.1》

Table 2.1.1 Top 11 engineering development fronts in agriculture

《Table 2.1.2》

Table 2.1.2 Annual number of core patents published for the Top 11 engineering development fronts in agriculture