A crop disease prevention and control method and system for improving the efficiency of pesticide use

Abstract

The present application relates to the technical field of crop disease prevention and control, more specifically, to a crop disease prevention and control method and system for improving pesticide use efficiency, soil samples and plant samples are collected, wherein the soil samples and plant samples come from different growth areas and environments, and different growth stages; the microbiome in the soil samples and plant samples is extracted, and the DNA of the microorganisms is extracted; the total DNA of the microorganisms is subjected to high-throughput metagenomic sequencing to obtain sequencing data; according to the sequencing data, a bioinformatics analysis method is used to identify pathogenic bacteria in the microbial community and drug-resistant genes existing in the pathogenic bacteria; a phylogenetic tree of the microbial community is constructed to analyze the succession relationship of the pathogenic bacteria; the pesticide resistance characteristics and change trend of the pathogen are analyzed; according to the analysis results of the microbial phylogenetic tree and the drug-resistant genes, disease prevention and control suggestions are generated to guide the selection and use of pesticides in the disease prevention and control process, greatly reducing the amount of pesticides used and improving the efficiency of disease prevention and control.

Description

Technical Field

[0001] This invention relates to the field of crop disease control technology, and more specifically, to a method and system for crop disease control that improves the efficiency of pesticide use. Background Technology

[0002] In recent years, with the intensification and scaling up of agricultural production, the importance of crop disease control in ensuring food security and sustainable agricultural development has become increasingly prominent. Traditional crop disease control methods mainly rely on fixed formulas and regular application of pesticides. While this method can control the spread of diseases to some extent, it also has many problems.

[0003] First, traditional methods lack precision. Because they don't take into account the actual drug resistance of pathogens, they often result in either excessive or insufficient pesticide use. Excessive use not only increases production costs but also causes serious environmental pollution; while insufficient use can lead to poor control and allow diseases to get out of control.

[0004] Secondly, this one-size-fits-all approach to pesticide application ignores the actual needs of different plots of land and crops at different growth stages, making it impossible to implement personalized control measures based on specific circumstances. This not only reduces the efficiency of pesticide use but may also accelerate the development and growth of pathogen resistance due to improper use.

[0005] Furthermore, traditional methods lack dynamic monitoring of changes in pathogen resistance. Over time, pathogens may develop new resistance mechanisms, and fixed treatment regimens cannot respond to these changes in a timely manner, leading to a gradual decline in control effectiveness.

[0006] Furthermore, while existing improved methods, such as molecular detection techniques based on single or a few genes, have improved detection accuracy to some extent, they still cannot fully reflect the complexity and dynamic changes of the microbial community. These methods often can only detect known resistance genes, making it difficult to promptly identify and address emerging resistance mechanisms.

[0007] Finally, current crop disease control generally lacks a deep understanding of the overall structure and function of the microbial community. The detection and control of single pathogens cannot reflect the complexity of the ecosystem and ignores the impact of intermicrobial interactions on disease occurrence and development.

[0008] In view of the above problems, there is an urgent need for a method and system for crop disease control that can comprehensively, dynamically, and accurately monitor pathogen resistance and guide the rational use of pesticides accordingly. This invention is an innovative solution proposed to address this need. Summary of the Invention

[0009] This invention proposes a method and system for crop disease control that improves pesticide application efficiency, aiming to solve technical problems such as insufficient precision of traditional control methods, inability to dynamically monitor changes in pesticide resistance, and difficulty in responding to emerging resistance mechanisms. By introducing microbial metagenomic analysis technology, this invention achieves comprehensive detection and accurate prediction of pesticide resistance in crop pathogens, thereby guiding the rational use of pesticides.

[0010] This invention provides a method for controlling crop diseases to improve the efficiency of pesticide application, comprising:

[0011] The sample acquisition steps include:

[0012] Soil and plant samples were collected, wherein the soil and plant samples came from different environments, different growth stages and different growth areas;

[0013] Total microbial DNA was extracted from the soil and plant samples.

[0014] Sequencing analysis steps include:

[0015] Based on the total DNA, high-throughput sequencing was performed to obtain sequencing data;

[0016] Based on the sequencing data, bioinformatics analysis methods were used to identify pathogens and their drug resistance genes;

[0017] Based on microbial metagenomic data, a microbial phylogenetic tree was constructed to analyze the succession relationships of pathogens;

[0018] Based on the identified drug resistance genes, the pesticide resistance characteristics and trends of pathogens were analyzed.

[0019] Output steps, including:

[0020] Based on the analysis results of the microbial growth tree and the drug resistance genes, disease control recommendations are generated, including the selection of pesticide types, dosages, and application frequencies.

[0021] Preferably, the sample acquisition step specifically includes:

[0022] Multiple sampling points were randomly selected within the designated plot of land;

[0023] Topsoil and deep soil were collected at each sampling point;

[0024] Leaves, stems, and roots of healthy plants and suspected diseased plants were collected.

[0025] Use sterile tools to collect samples and store them in sterile sampling bags.

[0026] Preferably, the high-throughput sequencing in the sequencing analysis specifically includes:

[0027] Sequencing libraries were constructed using a DNA library construction kit.

[0028] Sequencing was performed using paired-end sequencing technology;

[0029] Generate raw sequencing data.

[0030] Preferably, the bioinformatics analysis method in the sequencing analysis specifically includes:

[0031] The raw sequencing data were quality controlled using FASTQC software.

[0032] Use the cutadapt software to remove the connector sequence;

[0033] LEGO software was used for microbial species-level classification and calculation of drug resistance gene abundance.

[0034] Gene function was annotated using GO analysis.

[0035] Preferably, the sequencing analysis further includes:

[0036] Based on the analysis results of the drug resistance genes, the drug resistance index is calculated;

[0037] Based on the resistance index, the susceptibility of pathogens to different types of pesticides is determined.

[0038] Preferably, the process of constructing a microbial phylogenetic tree in the sequencing analysis includes:

[0039] The RAST procedure was used to select genes associated with the microorganism from the obtained gene abundance;

[0040] Gene homology analysis was performed using the protein orthologs mapper program;

[0041] Phylogenetic trees were constructed using the maximum likelihood method based on homologous genes.

[0042] Preferably, the output step further includes:

[0043] Generate a heatmap of drug resistance gene distribution to visually display the distribution of drug resistance genes in different pathogens;

[0044] Generate drug resistance trend reports to predict potential future drug resistance issues.

[0045] Preferably, a verification step is also included:

[0046] Based on the aforementioned pesticide use recommendations, pesticide application experiments were conducted in the experimental field;

[0047] Soil and plant samples were collected from the experimental field periodically, and the sample acquisition, sequencing analysis and output steps were repeated.

[0048] By comparing the changes in the distribution of resistance genes before and after the experiment, the effectiveness of pesticide use recommendations can be evaluated.

[0049] Preferably, a database update step is also included:

[0050] The results of the antibiotic resistance gene analysis, the microbial growth tree, and the pesticide use recommendations will be stored in a database.

[0051] Regularly update the database and establish a long-term monitoring system;

[0052] Based on the updated database, the pesticide application recommendation algorithm was optimized.

[0053] As preferred, including:

[0054] The sample collection module is used to collect soil and plant samples;

[0055] A DNA extraction module is used to extract total DNA from the soil and plant samples.

[0056] A sequencing module is used to perform high-throughput sequencing of the total DNA to obtain sequencing data;

[0057] The data analysis module is used for:

[0058] Based on the sequencing data, drug resistance genes of the pathogen were identified;

[0059] Construct a microbial phylogenetic tree to analyze the evolutionary relationships and drug resistance trends of pathogens;

[0060] The decision support module is used to generate pesticide application recommendations based on the analysis results of the resistance genes and the microbial growth tree.

[0061] The validation module is used to conduct pesticide application experiments in experimental fields to evaluate the effectiveness of pesticide use recommendations;

[0062] The database module is used to store and update the results of antibiotic resistance gene analysis, microbial growth trees, and pesticide application recommendations.

[0063] The beneficial effects of this invention are mainly reflected in the following aspects:

[0064] First, this invention significantly improves the accuracy and efficiency of crop disease control. Through comprehensive metagenomic analysis of soil and plant samples, this method can accurately identify and quantify various drug resistance genes, providing a scientific basis for developing personalized pesticide application plans. This not only improves disease control effectiveness but also significantly reduces pesticide usage, environmental pollution, and production costs.

[0065] Secondly, this invention enables dynamic monitoring and prediction of pathogen resistance. Through regular sampling and analysis, combined with a constructed microbial growth tree, this method can track the evolutionary trends of resistance genes and promptly identify emerging resistance mechanisms. This proactive monitoring allows pesticide application strategies to be adjusted in real time, effectively delaying the development of resistance.

[0066] Furthermore, the method of this invention has broad applicability and scalability. Although the embodiment uses rice bacterial blight as an example, the method is in principle applicable to various crop diseases. At the same time, with the continuous advancement of sequencing technology and bioinformatics analysis methods, there is still room for further improvement in the accuracy and efficiency of this method.

[0067] Furthermore, this invention promotes a deeper understanding of the interactions between crops, pathogens, and microbial communities. Through a comprehensive analysis of microbial community structure and function, this method provides new perspectives and data support for studying disease occurrence mechanisms and ecological control strategies.

[0068] Finally, the implementation of this invention helps promote the development of agricultural production towards precision, intelligence, and sustainability. By establishing a long-term monitoring system and a continuously optimized database, this method provides a scientific basis for formulating regional and long-term disease control strategies, which is of great significance for improving agricultural production efficiency and ensuring food security.

[0069] In summary, this invention innovatively solves several technical problems faced by traditional crop disease control methods by integrating multidisciplinary technologies. It shows significant advantages in improving control effectiveness, reducing environmental impact, and delaying the development of drug resistance, providing a new, efficient, precise, and sustainable approach to disease control for modern agricultural production. Attached Figure Description

[0070] Figure 1 This is a flowchart of the method of the present invention.

[0071] Figure 2 This is a logic block diagram of the data analysis module of the present invention.

[0072] Figure 3 This is a logical block diagram of the decision support module of the present invention. Detailed Implementation

[0073] Please refer to Figure 1-3This invention provides a method and system for crop disease control that improves pesticide application efficiency. This method utilizes microbial metagenomic analysis technology to achieve precise detection and prediction of pesticide resistance in crop pathogens, thereby guiding the rational use of pesticides, improving control effectiveness, and reducing environmental pollution risks.

[0074] Specifically, the method of the present invention includes the following steps:

[0075] First, in the sample acquisition step, this method collects soil and plant samples. These samples come from different environments, growth stages, and growth regions to ensure representativeness and diversity. For example, samples can be collected before sowing, during the seedling stage, flowering stage, and fruiting stage, while considering differences in field location and soil depth. Total microbial DNA is extracted from the soil and plant samples; based on the total DNA, high-throughput sequencing is performed to obtain sequencing data; according to the sequencing data, bioinformatics analysis methods are used to identify pathogens and their drug resistance genes.

[0076] In the sequencing analysis, this method first performs high-throughput sequencing on the extracted total microbial DNA. High-throughput sequencing technology can rapidly and massively acquire genetic information from samples. Preferably, the Illumina sequencing platform can be used for paired-end sequencing, with a read length of 150 bp and a sequencing depth of at least 10 Gb to ensure data quality and coverage.

[0077] After obtaining the sequencing data, this method uses bioinformatics analysis to identify drug resistance genes in pathogens. This step is one of the core aspects of this invention. Specifically, the ARIBA (Antimicrobial Resistance Identification By Assembly) software can be used to identify and classify drug resistance genes. ARIBA software can simultaneously perform sequence assembly and drug resistance gene annotation, improving the accuracy and efficiency of the analysis.

[0078] After identifying drug-resistant genes, this method constructs a microbial phylogenetic tree based on this data to analyze the evolutionary relationships and drug resistance trends of pathogens. During the phylogenetic tree construction, the RAxML (Randomized Axelerated Maximum Likelihood) software is used, employing the maximum likelihood method. This step helps to understand the structure and dynamic changes of pathogenic communities, providing important evidence for formulating control strategies.

[0079] In the output step, this method generates pesticide use recommendations based on the analysis results of resistance genes and the microbial growth tree. These recommendations include pesticide type, dosage, and application time. For example, if a significant increase in the abundance of β-lactam antibiotic resistance genes is detected, it may be necessary to reduce the use of the relevant pesticide and instead choose pesticides with different mechanisms of action.

[0080] In one embodiment of the method of the present invention, the sample acquisition steps can be further refined. Specifically, multiple sampling points can be randomly selected within a predetermined plot of land. The number of sampling points can be determined according to the size of the plot, typically 3-5 sampling points per hectare. At each sampling point, topsoil (0-20cm) and deep soil (20-40cm) are collected. Simultaneously, leaves, stems, and roots of healthy plants and suspected diseased plants are collected. During the sampling process, sterile tools such as sterilized shovels and scissors are used, and the samples are immediately placed in sterilized self-sealing bags for storage and transportation at 4°C.

[0081] In sequencing analysis, the specific workflow of high-throughput sequencing deserves a detailed explanation. First, a sequencing library is constructed using a DNA library construction kit (such as the Illumina TruSeq DNA PCR-Free Library Preparation Kit). The library construction process includes DNA fragmentation, end repair, A-tail addition, adapter ligation, and PCR amplification. Then, paired-end sequencing is performed, typically using the Illumina NovaSeq 6000 system with a read length of 2 × 150 bp. After sequencing, raw sequencing data is generated, usually in FastQ format.

[0082] This invention utilizes a multidisciplinary approach to achieve precise analysis and prediction of crop pathogen resistance. This method not only improves the targeting and effectiveness of pesticide application but also helps reduce unnecessary pesticide use, thereby mitigating environmental pollution risks and promoting sustainable agricultural development. Furthermore, through long-term monitoring and data accumulation, this method can provide valuable data support for research on the evolution of pathogen resistance, advancing scientific research in related fields.

[0083] In a preferred embodiment of the present invention, the bioinformatics analysis method in sequencing analysis is further refined to improve the accuracy and efficiency of data processing. Specifically, the method first uses FASTQC software to perform quality control on the raw sequencing data. FASTQC can generate a series of charts and statistics to help researchers quickly assess the quality of sequencing data. For example, it can check parameters such as the sequencing quality score of each base, GC content distribution, and sequence repetition level. Typically, it is expected that the average quality score of each base is not less than 30 (i.e., Q30), which means that the sequencing error rate is less than 0.1%. This can effectively reduce the impact of low-quality data on subsequent analysis and improve the reliability of the overall data. FASTQC can quickly identify potential problems, such as GC content deviation and sequence repetition, helping researchers adjust experimental design or sequencing parameters in a timely manner and avoid wasting time and resources.

[0084] After quality control, this method uses the cutadapt software to remove adapter sequences. During sequencing, adapter sequences may be read by the sequencer, and these sequences need to be removed to avoid affecting subsequent analysis. The cutadapt software can efficiently identify and remove these adapter sequences. Preferably, a minimum overlap length of 3 bp and an allowable mismatch rate of 10% can be set to balance removal efficiency and data retention.

[0085] Cutadapt identifies and removes known adapter sequences by comparing them with sequencing reads. It uses a dynamic programming algorithm to find the optimal match and performs cuts based on a set minimum overlap length and maximum mismatch rate. This algorithm maximizes the retention of valid data while maintaining high removal efficiency. By setting reasonable minimum overlap lengths (3 bp) and mismatch rates (10%), it avoids erroneous cuts of valid data while removing adapter sequences, ensuring the accuracy of subsequent analyses. The presence of adapter sequences can interfere with subsequent gene annotation and functional analysis. Efficient adapter sequence removal can significantly improve the quality of downstream analyses and reduce false positive results.

[0086] Next, the method of this invention uses LEGO (LEt's GO) software for microbial species-level classification and antimicrobial resistance gene abundance calculation. LEGO software is a tool specifically designed for metagenomic data analysis; it employs an improved nearest neighbor classification algorithm, enabling rapid and accurate classification of microorganisms. For species classification, a similarity threshold of 97% can be set, a commonly used standard for distinguishing bacterial species. For antimicrobial resistance gene abundance calculation, LEGO employs a k-mer-based method, which can effectively handle large-scale data.

[0087] LEGO software employs an improved nearest neighbor classification algorithm, calculating the similarity between sequencing reads and known sequences in a reference database. A similarity threshold of 97% is typically used, a common standard for distinguishing bacterial species. For calculating the abundance of antibiotic resistance genes, LEGO uses a k-mer-based method, estimating the relative abundance of resistance genes by statistically analyzing the frequency of each k-mer across different samples. LEGO's improved algorithm enables fast and accurate classification of microorganisms on large-scale datasets, significantly improving classification efficiency, especially when dealing with complex metagenomic data. The k-mer-based method effectively handles large-scale data, providing accurate estimates of antibiotic resistance gene abundance and helping researchers understand the distribution of resistance genes in different samples.

[0088] Finally, this method uses Gene Ontology (GO) analysis to annotate gene function. GO analysis can describe gene function from three aspects: molecular function, biological process, and cellular components. Common tools such as Blast2GO or DAVID can be used for GO analysis. Preferably, the E-value threshold can be set to 1e. -5 This ensures the reliability of annotations and avoids mis-annotating low-quality alignment results. GO analysis can describe gene function from three aspects: molecular function, biological process, and cellular components, helping researchers to more comprehensively understand the mechanisms of gene action and guiding subsequent experimental design and data analysis.

[0089] In another embodiment of the invention, the sequencing analysis further includes calculating a resistance index based on the analysis results of resistance genes. This step provides a quantitative indicator for assessing the overall drug resistance level of the pathogen. The resistance index (RI) can be calculated using the following formula:

[0090]

[0091] Where n is the number of detected drug resistance gene types, The weight of the i-th resistance gene can be determined based on its importance; for example, some resistance genes may exhibit stronger resistance to specific pesticides and therefore should be assigned a higher weight. By standardizing the abundance of each resistance gene to a proportion relative to the highest abundance, the influence of absolute abundance differences between different samples can be eliminated, making the RI more comparable. A i Let A be the abundance of the i-th drug resistance gene. max This is the highest abundance value among all drug resistance genes.

[0092] After calculating the resistance index, this method determines the susceptibility of pathogens to different types of pesticides based on this index. For example, the RI value can be divided into the following intervals: RI < 0.3 indicates low resistance, 0.3 ≤ RI < 0.6 indicates moderate resistance, and RI ≥ 0.6 indicates high resistance. This classification can intuitively guide the selection and use of pesticides.

[0093] In constructing the microbial phylogenetic tree, the method of this invention employs a series of bioinformatics tools and algorithms. First, the RAST (Rapid Annotation using Subsystem Technology) program is used to select microbial-related genes from the obtained gene abundance. RAST is an automated genome annotation system capable of rapidly identifying and annotating genes, especially those related to metabolism.

[0094] Subsequently, this method utilizes the protein orthologs mapper program for gene homology analysis. This step aims to identify homologous genes among different species, laying the foundation for constructing a phylogenetic tree. When performing homology analysis, the E-value threshold can be set to 1e. -10 This is to ensure a highly reliable homology.

[0095] Finally, based on homologous genes, this method constructs a phylogenetic tree using the maximum likelihood method. Maximum likelihood is a statistical method that infers the most probable evolutionary history from observed data. In practice, this process can be implemented using the RAxML software. RAxML employs a fast algorithm capable of handling large-scale datasets. Preferably, the GTR+Γ model can be chosen as the nucleotide substitution model; this is a widely used complex model that can effectively describe the evolutionary process of DNA sequences.

[0096] Maximum likelihood estimation is a statistical method that identifies the evolutionary history that best explains observed data. The RAxML software uses a fast algorithm to calculate the maximum likelihood value and select the most probable phylogenetic tree. The GTR+Γ model is a widely used complex model that effectively describes the evolutionary process of DNA sequences, particularly considering the differences in evolutionary rates at different sites (modeled using Gamma distribution). By constructing phylogenetic trees, the evolutionary relationships between different pathogens can be revealed, helping researchers understand the transmission pathways of antibiotic resistance across different species. Based on the results of phylogenetic trees, more targeted control strategies can be developed, such as selecting pesticides effective against specific pathogens or developing new resistant varieties.

[0097] The output steps of this invention not only generate pesticide application recommendations but also include generating a heatmap of resistance gene distribution and a resistance trend report. The heatmap visually displays the distribution of resistance genes in different pathogens, helping researchers quickly identify major resistance gene types and distribution patterns. When generating the heatmap, the pheatmap package in R can be used, with an appropriate color scheme selected (e.g., blue to red representing abundance from low to high), and hierarchical clustering performed to reveal the relationship between samples and genes.

[0098] The Antimicrobial Resistance Trends Report predicts potential future resistance issues. This report uses time series analysis methods for forecasting, based on historical data and current analysis results. For example, an ARIMA (Autoregressive Integrated Moving Average) model can be used for trend analysis. When selecting model parameters, the AIC (Akaike Information Criterion) can be used to balance model complexity and goodness of fit.

[0099] Through these detailed analysis and visualization steps, the method of this invention not only provides a comprehensive picture of the current state of pesticide resistance but also predicts future trends, offering a strong scientific basis for the formulation and adjustment of pesticide application strategies. This data-driven approach significantly improves the accuracy and foresight of crop disease control, and is expected to significantly improve pesticide application efficiency, reduce unnecessary pesticide use, thereby reducing environmental pressure and promoting sustainable agricultural development.

[0100] The method of this invention also includes a crucial validation step in practical applications to ensure the effectiveness and reliability of the pesticide application recommendations. In this validation step, pesticide application experiments are first conducted in experimental fields based on the pesticide application recommendations derived from the aforementioned analysis. The experimental fields are typically set up using a randomized block design to eliminate the influence of differences in field environments. Preferably, 3-5 replicates can be set up, with each plot area not less than 20 square meters, to ensure the statistical significance of the experimental results.

[0101] During the experiment, this method requires the regular collection of soil and plant samples from the experimental field. The sampling frequency can be determined based on the crop growth cycle and disease occurrence patterns, and sampling during the main growth stages (such as seedling, flowering, and fruit development stages) is generally appropriate. The collected samples are then processed using the same procedures as the initial analysis, i.e., repeating the aforementioned sample acquisition, sequencing analysis, and output steps. This periodic sampling and analysis allows researchers to dynamically monitor changes in drug resistance genes.

[0102] The key to validation lies in comparing the changes in the distribution of resistance genes before and after the experiment, thereby assessing the effectiveness of pesticide use recommendations. This comparison can be quantified by calculating the rate of change in the abundance of resistance genes:

[0103]

[0104] Preferably, a change rate less than -20% can be considered a significant decrease, between -20% and 20% can be considered stable, and greater than 20% can be considered a significant increase. This quantitative analysis can objectively evaluate the effectiveness of pesticide application strategies and provide a basis for further optimization.

[0105] Another innovation of this invention lies in the establishment of a dynamically updated database system. This database update step is crucial for the long-term effectiveness of the method. First, this method stores key information such as the results of antibiotic resistance gene analysis, microbial growth trees, and pesticide application recommendations into a specially designed database. This database employs a relational database management system (such as PostgreSQL) to ensure data integrity and consistency.

[0106] Regular database updates are crucial for maintaining system timeliness. Ideally, a comprehensive update can be performed quarterly, with an automatic update mechanism set up to update relevant fields as new analysis results are generated. This update mechanism ensures that the database always contains the latest drug resistance information and prevention and control recommendations.

[0107] By establishing such a long-term monitoring system, this method can not only track the changing trends of drug resistance in individual fields or regions, but also analyze the evolution of drug resistance on a larger spatiotemporal scale. For example, time series analysis methods, such as the seasonal ARIMA model, can be used to predict future drug resistance trends. Such predictions have important guiding significance for formulating long-term disease control strategies.

[0108] Based on a continuously updated database, the method of this invention also includes continuous optimization of the pesticide application recommendation algorithm. This optimization process employs machine learning methods, such as random forests or gradient boosting decision trees, to continuously learn new data patterns and improve prediction accuracy. After each optimization, cross-validation is performed to ensure that the new algorithm maintains good performance under various conditions.

[0109] Finally, the present invention also provides a crop disease control system corresponding to the above-described method. This system consists of multiple functional modules, each responsible for a specific step in the method.

[0110] Sample collection module 1 is responsible for on-site sampling. This module is equipped with specialized sampling tools, such as soil samplers and plant tissue samplers, and has a built-in GPS positioning system to ensure accurate recording of sampling locations. Preferably, this module also includes a mobile application for recording environmental parameters (such as temperature and humidity) and plant conditions during sampling.

[0111] DNA extraction module 2 employs automated extraction equipment, such as a magnetic bead nucleic acid extractor, capable of efficiently extracting total DNA from various samples. The module's design takes into account the characteristics of different sample types, automatically adjusting extraction parameters based on sample properties to ensure both DNA quality and yield.

[0112] Sequencing module 3 employs the latest high-throughput sequencing technologies, such as the Illumina NovaSeq system, enabling the sequencing of a large number of samples in a short time. This module also includes an automated library preparation system, significantly reducing the need for manual operations.

[0113] Data analysis module 4 is the core of the system, integrating various bioinformatics analysis tools and algorithms. This module employs a distributed computing architecture, enabling rapid processing of massive sequencing data. Preferably, this module also possesses machine learning capabilities, allowing for continuous optimization of analysis algorithms as data accumulates.

[0114] In data analysis module 4, a novel method, Dynamic Resistance Evaluation and Microbial Community Structure Analysis (DREMCSA), is introduced.

[0115] First, construct a drug resistance gene abundance matrix R, where r ij This represents the relative abundance of the j-th drug resistance gene in the i-th sample.

[0116]

[0117] Where m is the number of samples and n is the number of detected drug resistance genes. In the preferred embodiment, the LEGO software can be used as the specific tool for constructing the matrix.

[0118] The drug resistance gene abundance matrix R is an m×n matrix. This matrix is ​​constructed based on metagenomic sequencing data, and its abundance is determined by calculating the number of sequence reads of different drug resistance genes in each sample.

[0119] By constructing a drug resistance gene abundance matrix, the distribution of drug resistance genes in different samples can be accurately identified. This helps to understand which drug resistance genes are more common in specific environments, thus providing a basis for subsequent analysis and decision-making.

[0120] Each row in the matrix represents a sample, and each column represents a drug resistance gene. By performing statistical analysis on the matrix (such as principal component analysis and cluster analysis), the differences in microbial community structure among different samples can be assessed, revealing potential trends in drug resistance changes.

[0121] The matrix can be visually displayed in the form of heatmaps, helping researchers quickly identify the main types and distribution patterns of drug resistance genes, and guiding subsequent experimental design and data analysis.

[0122] A topological drug resistance network is constructed based on the drug resistance gene abundance matrix. The association strength s between drug resistance genes is defined. ij for:

[0123]

[0124] in, and , respectively, represent the average abundance of the i-th and j-th drug resistance genes.

[0125] The strength of associations between drug resistance genes ij The Pearson correlation coefficient measures the co-expression relationship of two drug resistance genes across different samples. The numerator represents the covariance of the two genes across all samples, while the denominator is the product of their respective standard deviations, ensuring the standardization of the results. The Pearson correlation coefficient can be used to assess co-expression relationships between genes, while phylogenetic trees can reveal evolutionary relationships between genes. Using both together can provide a more comprehensive analysis of drug resistance gene networks.

[0126] By constructing topological resistance networks, the complex interactions between resistance genes can be captured. For example, some resistance genes may work synergistically, leading to stronger resistance, while others may inhibit each other. This network structure helps in understanding the mechanisms of resistance propagation. Within the network, certain nodes (i.e., resistance genes) may possess high centrality or betweenness, indicating their crucial role in resistance propagation. Identifying these key nodes can help researchers develop targeted control strategies to reduce the spread of resistance. The structure of resistance networks also changes over time and with environmental variations. By regularly updating the network, interactions between resistance genes can be monitored in real time, allowing for timely adjustments to control measures and preventing further development of resistance.

[0127] Then, using concepts from group theory, a dynamic drug resistance index (DRI) is defined. First, the topological network is considered as a group G, and the operations on the group are defined as the synergistic effects between drug-resistant genes. Let φ be a homomorphic mapping from group G to the set of real numbers R, representing the drug resistance strength. Then, the DRI is defined as:

[0128] DRI = ∑ g∈G φ(g)·∏ h∈N(g) (1+s gh ),

[0129] Where N(g) is the set of adjacent nodes of the drug resistance gene g in the topological network, s gh The correlation strength between g and h.

[0130] This formula defines a dynamic resistance index (DRI), which is constructed based on group theory concepts. Group G represents the set of nodes in the resistance gene network, φ(g) is a homomorphism representing the resistance strength of each resistance gene, and N(g) is the set of neighboring nodes of node g. gh It represents the association strength between nodes g and h. The product term ∏ in the formula... h∈N(g) (1+s gh () indicates the synergistic effect between drug resistance genes.

[0131] By incorporating group theory, the DRI (Drug Resistance Index) considers not only the abundance of individual resistance genes but also their interactions. This allows the DRI to more comprehensively reflect the resistance level of the entire system, providing a scientific basis for developing integrated pest management strategies. Over time, the structure of the resistance gene network may change, causing fluctuations in the DRI. By monitoring the long-term trend of DRI changes, potential future resistance problems can be predicted, allowing for proactive countermeasures. The DRI can serve as an indicator for assessing the resistance risk in different fields or regions. Based on the DRI level, pesticides and other control resources can be allocated rationally, avoiding unnecessary waste and improving control efficiency. The DRI focuses more on the synergistic effects between resistance genes, while the RI emphasizes the abundance and importance of individual resistance genes. Combining these two indices can provide a more comprehensive assessment of resistance.

[0132] Next, the characteristic function χ(t) of the microbial community structure is defined:

[0133]

[0134] Where F(x) is the microbial species abundance distribution function, and t is a complex parameter. By analyzing the properties of χ(t), the diversity and evenness of the microbial community can be assessed.

[0135] This formula defines the characteristic function χ(t) of microbial community structure, which is the Fourier transform of the microbial species abundance distribution function F(x). The characteristic function is a commonly used tool in probability theory, capable of describing the distribution characteristics of random variables from a frequency domain perspective. By analyzing the properties of χ(t), the diversity and evenness of the microbial community can be assessed.

[0136] The characteristic function χ(t) can be used to assess the diversity of microbial communities. For example, if χ(t) shows a large amplitude in the high-frequency range, it indicates the presence of many rare species in the community; while a large amplitude in the low-frequency range indicates a concentration of dominant species. This assessment method can help researchers understand the health of farmland ecosystems. By comparing the characteristic function at different time points, the dynamic changes of the microbial community can be analyzed. For example, if the characteristic function remains stable over a period of time, it indicates a relatively stable community structure; conversely, if the characteristic function changes significantly, it may indicate that the ecosystem has been disturbed by external factors, requiring corresponding protective measures. The structure of a microbial community is closely related to its function. By analyzing the characteristic function, the functional roles of different microorganisms in the community can be inferred, thereby predicting their impact on crop growth and disease control. For example, some microorganisms may promote plant growth, while others may enhance crop disease resistance. Characteristic functions can be used to assess community diversity and evenness, while heatmaps and cluster analysis can visually display changes in community structure. Combining both can provide a more comprehensive assessment of the microbial community.

[0137] Finally, a number theory-based model for predicting drug resistance evolution is introduced. The drug resistance evolution function E(n) is defined as follows:

[0138] E(n)=∑ d|n μ(d)·DRI n / d ,

[0139] Where n is the time step, μ(d) is the Möbius function, and d|n means that d divides n. DRI is the previously calculated dynamic resistance index.

[0140] This model utilizes the Möbius inversion formula in number theory to capture the periodic and mutational characteristics of drug resistance evolution.

[0141] This formula defines the drug resistance evolution function E(n), which is constructed based on the Möbius inversion formula in number theory. The Möbius function μ(d) is an important number-theoretic function used to handle divisibility relations. The summation term ∑ in the formula... d|n μ(d)·DRI n / d This represents the weighted average of the drug resistance index (DRI) at time step n. The Möbius inversion formula can capture the periodic and abrupt characteristics of drug resistance evolution.

[0142] By introducing the Möbius function from number theory, E(n) can capture the long-term trend of pesticide resistance evolution. For example, some resistant genes may exhibit periodic fluctuations within specific time intervals, while other genes may undergo mutations. This predictive ability helps in developing long-term disease control strategies and preventing potential risks in advance. The Möbius inversion formula is particularly suitable for handling discrete events (such as gene mutations). When the resistance index DRI mutates, the value of E(n) changes significantly. By monitoring changes in E(n), resistance mutation events can be detected in a timely manner, allowing for emergency response measures. Based on the predicted results of E(n), future pesticide usage and application timing can be rationally planned. For example, if a significant increase in resistance is predicted within a certain period, more alternative pesticides can be stockpiled in advance to avoid control failure due to rising resistance. The Möbius inversion formula can capture the periodicity and mutation characteristics of pesticide resistance evolution, while the ARIMA model is more suitable for handling long-term trend predictions of time series data. Combining these two methods can provide more accurate pesticide resistance predictions.

[0143] The DREMCSA method enables a comprehensive analysis, from quantifying the abundance of drug resistance genes to predicting the evolution of drug resistance. This method not only considers the abundance of individual drug resistance genes but also captures the interactions between them through topological networks. It quantifies the overall drug resistance level using group theory and achieves in-depth analysis of microbial community structure and drug resistance evolution through characteristic functions and number theory methods.

[0144] The innovation of this comprehensive approach is mainly reflected in:

[0145] 1. Using topological networks to characterize the complex interactions between drug resistance genes.

[0146] 2. By introducing the concept of group theory, the drug resistance gene network is regarded as an algebraic structure, which better describes the holistic nature of the system.

[0147] 3. Using characteristic functions to analyze microbial community structure provides a new method for diversity assessment.

[0148] 4. By constructing an evolutionary model of drug resistance using the Möbius function in number theory, it is possible to predict the long-term evolutionary trend of drug resistance.

[0149] The application of this method will significantly improve the accuracy and foresight of crop disease control, and provide a more reliable scientific basis for the formulation of pesticide application strategies.

[0150] Decision support module 5 generates specific pesticide application recommendations based on data analysis results. This module employs an expert system design, combining pesticide resistance analysis results with an agronomic knowledge base to provide personalized control solutions.

[0151] Verification module 6 is responsible for monitoring the experimental field, including automated sampling and data acquisition systems. This module can monitor various indicators of the experimental field in real time and promptly detect anomalies.

[0152] Database module 7 employs distributed storage technology to ensure secure storage and fast access to large-scale data. This module also features data mining capabilities, enabling the discovery of valuable patterns and trends from historical data.

[0153] Through the collaborative work of these modules, the system of this invention achieves full automation from sample collection to decision support, greatly improving the efficiency and accuracy of crop disease control. The modular design of the system also makes future upgrades and expansions simple and easy, laying the foundation for intelligent and precise agricultural production.

[0154] To verify the effectiveness of the crop disease control method and system proposed in this invention, researchers conducted a field trial for one growing season at an agricultural research base. The trial selected rice bacterial blight, a common local disease caused by Xanthomonas oryzaepv. oryzae, as the research subject. This disease is one of the major bacterial diseases in rice production.

[0155] Example 1 illustrates disease control using the method of this invention. First, soil and plant samples were collected before rice planting, during the tillering stage, and at the heading stage, respectively, for microbial metagenomic sequencing and analysis. Based on the analysis results, researchers identified several major resistance genes, including qnrS (quinolone antibiotic resistance gene) and bla_TEM (β-lactam antibiotic resistance gene). Based on the distribution and abundance of these resistance genes, targeted pesticide application recommendations were systematically generated, including reducing the frequency of quinolone pesticide use and increasing the proportion of biological agents used.

[0156] Comparative Example 1 used the traditional fixed-formula application method, that is, spraying pesticides regularly according to the conventional recommended dosage, without considering the changes in the pathogen's resistance.

[0157] Both experiments were conducted with three replicates, each plot measuring 50 square meters. During the experiments, researchers regularly observed and recorded disease occurrence and measured yield at rice maturity. Simultaneously, samples were collected periodically throughout the growing season to monitor changes in antibiotic resistance genes.

[0158] The main test indicators and their experimental methods are as follows:

[0159] 1. Disease incidence rate: Surveys were conducted every 7 days. 100 rice plants were randomly selected, the number of infected plants was recorded, and the incidence rate was calculated.

[0160] 2. Disease index: A scale of 0-9 is used, where level 0 indicates no disease and level 9 indicates the entire plant is dead. The disease index is calculated every 7 days.

[0161] 3. Pesticide usage: Record the total amount of pesticides used throughout the growing season.

[0162] 4. Abundance of drug resistance genes: The relative abundance of qnrS and bla_TEM genes was determined by real-time quantitative PCR.

[0163] 5. Rice yield: Harvest at maturity and measure the actual yield.

[0164] 6. Economic benefits: Calculate the input-output ratio.

[0165] The experimental results are shown in the table below:

[0166] index Example 1 Comparative Example 1 Final disease incidence 12.3％ 18.7％ Final disease index 15.6 24.2 Pesticide application rate (kg / ha) 2.8 4.5 relative abundance of qnrS gene 0.85 2.37 relative abundance of bla_TEM genes 1.12 2.85 Rice yield (t / ha) 7.8 7.2 Economic benefits (input-output ratio) 1:3.2 1:2.6

[0167] The experimental results show that the method of the present invention exhibits significant advantages in several aspects. First, in terms of disease control, the final disease incidence rate and disease index of Example 1 using the method of the present invention are significantly lower than those of Comparative Example 1, decreasing by 34.2% and 35.5%, respectively. This indicates that the precise application strategy of the present invention can more effectively suppress the development of diseases.

[0168] Secondly, regarding pesticide usage, Example 1 reduced it by 37.8% compared to Comparative Example 1, which not only lowered production costs but also reduced negative environmental impacts. More importantly, the relative abundance of resistance genes in Example 1 was significantly lower than in Comparative Example 1. The relative abundance of qnrS and bla_TEM genes decreased by 64.1% and 60.7%, respectively, indicating that the method of this invention can effectively inhibit the development of resistance and lay the foundation for long-term sustainable control.

[0169] In terms of yield, although Example 1 using the method of the present invention used less pesticide, its rice yield was 8.3% higher than that of Comparative Example 1. This may be due to the more precise disease control reducing the impact of diseases on rice growth, and may also be related to the improved ecosystem balance resulting from reduced pesticide use.

[0170] Finally, from an economic perspective, the input-output ratio of Example 1 is 1:3.2, which is significantly higher than that of Comparative Example 1 (1:2.6). This indicates that the method of the present invention is not only technically superior but also more economically attractive.

[0171] These results fully demonstrate the superiority of the crop disease control method based on microbial metagenomic analysis proposed in this invention. This method, through precise identification and monitoring of drug-resistant genes, achieves precise and personalized pesticide application, not only improving disease control effectiveness but also significantly reducing pesticide usage, inhibiting the development of drug resistance, and simultaneously increasing crop yield and economic benefits. This method provides a new approach and effective means to address the current problems in agricultural production, such as low disease control efficiency, excessive pesticide use, and increasingly serious drug resistance.

[0172] It should be noted that the advantages of the method of this invention may become more apparent over time. Long-term follow-up studies may reveal that drug resistance problems will worsen under traditional methods, while the method of this invention may maintain good control effects. Therefore, it is recommended to conduct longer-term observations in future studies to comprehensively evaluate the long-term benefits of this method.

[0173] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, substitutions, or improvements made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A crop disease prevention and control method for improving the efficiency of pesticide use, characterized in that, The method comprises the following steps: a sample acquisition step, comprising: collecting soil samples and plant samples, wherein the soil samples and plant samples are from different environments, different growth stages and different growth areas; extracting total microbial DNA from the soil samples and plant samples; a sequencing analysis step, comprising: based on the total DNA, performing high-throughput sequencing to obtain sequencing data; using bioinformatics analysis methods to identify pathogenic bacteria and their drug resistance genes according to the sequencing data; constructing a microbial phylogenetic tree based on the microbial metagenomic data to analyze the succession relationship of pathogenic bacteria; analyzing the pesticide resistance characteristics and change trend of pathogens based on the identified drug resistance genes; and an output step, comprising: generating disease prevention and control suggestions based on the microbial phylogenetic tree and the analysis results of the drug resistance genes, wherein the disease prevention and control suggestions include the selection of pesticide types, the use dose and the application frequency; The bioinformatics analysis method in the sequencing analysis specifically comprises: using FASTQC software to perform quality control on the raw sequencing data; using cutadapt software to remove adapter sequences; using LEGO software to perform microbial species level classification and drug resistance gene abundance calculation; and using GO analysis to annotate gene functions; The sequencing analysis further comprises: calculating a drug resistance index based on the analysis results of the drug resistance genes; and determining the sensitivity of pathogenic bacteria to different types of pesticides according to the drug resistance index; The process of constructing a microbial phylogenetic tree in the sequencing analysis comprises: using RAST program to select genes related to the microorganism from the obtained gene abundance; using proteinorthologsmapper program to perform gene homology analysis; and using maximum likelihood method to construct a phylogenetic tree based on homologous genes; The output step further comprises: generating a drug resistance gene distribution heat map to visually display the drug resistance gene distribution of different pathogenic bacteria; and generating a drug resistance trend report; The drug resistance index RI is calculated by the following formula: ， wherein n is the number of detected drug resistance gene types, is the weight of the ith drug resistance gene, is the abundance of the ith drug resistance gene, is the value of the highest abundance among all drug resistance genes; The method further comprises a verification step: based on the pesticide use suggestions, performing pesticide application experiments in a test field; periodically collecting soil and plant samples from the test field, and repeating the sample acquisition step, the sequencing analysis and the output step; comparing the changes in drug resistance gene distribution before and after the experiment to evaluate the effectiveness of the pesticide use suggestions; The method further comprises a database updating step: storing the drug resistance gene analysis results, the microbial phylogenetic tree and the pesticide use suggestions in a database; periodically updating the database to establish a long-term monitoring system; and optimizing the pesticide use suggestion algorithm based on the updated database.

2. The method of claim 1, wherein, The sample acquisition step specifically comprises: randomly selecting multiple sampling points in a predetermined plot; collecting surface soil and deep soil at each sampling point; collecting leaves, stems and roots of healthy plants and suspected diseased plants; using sterile tools to collect samples and storing the samples in sterilized sampling bags.

3. The method of claim 1, wherein, The high-throughput sequencing in the sequencing analysis specifically comprises: using a DNA library construction kit to construct a sequencing library; using a double-end sequencing technology to perform sequencing; and generating raw sequencing data.

Images