Effective disease resistance gene prediction method based on rice blast fungal communities in different rice-growing areas

Abstract

This invention relates to the field of bioinformatics, specifically to a method for predicting effective resistance genes based on rice blast pathogens from different rice-growing regions. This method analyzes different rice varieties in different rice-growing regions, analyzing their infectivity by examining infection with strains containing different abnormal avirulent gene types. It combines historical data to analyze the extent of disease spread in the respective rice-growing regions by all strains corresponding to these abnormal avirulent genes and their growth adaptability in different regions to determine the strain's harmfulness. Based on the phylogenetic relationships between different strains and their corresponding rice plants, it analyzes the strain's damage tolerance. Furthermore, it adjusts the attention weights of the nodes corresponding to each strain based on the obtained damage tolerance, resulting in a graph neural network prediction. This invention analyzes the degree of damage caused by each strain to rice in rice plants infected with rice blast in different rice-growing regions, and predicts the effective resistance of different rice varieties to these strains based on their stable expansion ability, thus improving the accuracy of resistance gene effectiveness prediction.

Description

Technical Field

[0001] This invention relates to the field of bioinformatics, specifically to an effective method for predicting disease resistance genes based on rice blast pathogens from different rice-growing regions. Background Technology

[0002] Currently, when predicting effective resistance genes of rice blast fungal communities in different rice-growing areas based on bioinformatics, rice blast fungal strains from various rice-growing areas are usually collected first. Molecular detection or gene sequencing is then performed on the screened strains to obtain raw genomic data. Bioinformatics tools (such as gene annotation and sequence alignment) are then used to analyze and obtain the avirulence genotype of each strain, and the variation type (such as mutation and deletion) of the corresponding avirulence gene of each strain is extracted. The obtained abnormal information and the specific lesion manifestations of the corresponding rice blast fungal rice plants are then input into neural networks (such as CNN, GNN, etc.), so as to predict and evaluate the effectiveness of each resistance gene based on the genotype-phenotype association of rice plants.

[0003] However, in actual testing, there is a certain biological correlation between rice blast fungus and its damage to rice. Existing methods, which only rely on variation data of avirulent gene profiles to predict the effectiveness of resistance genes, cannot accurately reflect the relationship between abnormal genes of each rice blast fungus strain and its actual phenotype (i.e., pathogenicity). For example, a mutated avirulent gene may be used as an early warning indicator, but its protein product may have a similar function during infection, making the resistance gene still effective, resulting in a high margin of error. Summary of the Invention

[0004] To address the technical problems in the prior art, the present invention aims to provide an effective method for predicting disease resistance genes based on different rice blast pathogen groups in different rice-growing regions. The specific technical solution adopted is as follows: This invention provides an effective method for predicting resistance genes based on rice blast fungal communities in different rice-growing areas, the method comprising: In different rice-growing areas, the growth stage of each rice variety and the ratio of lesion area of ​​infected rice plants were obtained at each sampling time; and the abnormal non-virulence gene type of each strain in the infected rice plants at each sampling time was determined. For each infected rice plant of each variety in each rice-growing area in each sampling, the expansion index of each abnormal nonvirulent gene type of each strain was determined based on the existence duration of each abnormal nonvirulent gene type of each strain and the increase in the number of infected rice plants, combined with the increase in the lesion area ratio. For each rice variety in each rice-growing area under the current sampling, the growth impact index of each strain on each rice variety in each rice-growing area is obtained based on the deviation distribution of the expansion index of each strain corresponding to all abnormal non-toxic gene types from the historical situation. Based on the similarity of environmental data between each rice-growing area and other rice-growing areas, the deviation distribution of each strain on each rice variety in the growth impact index is analyzed to determine the hazard capacity index of each strain. Under the current sampling conditions, the highly transmissible strains of each rice variety are identified based on the hazard capacity index. For each highly transmissible strain of each rice variety, the transmission chain of each highly transmissible strain is determined based on its phylogenetic relationship with other rice varieties and the similarity of its hazard capacity index in each growth stage cycle. The hazard tolerance of each strain is determined by the distribution of its hazard capacity index on the maximum transmission chain. Predictions are made by adjusting the attention weights of the current strain in the graph neural network based on hazard durability.

[0005] Furthermore, the method for obtaining the expansion index includes: For any abnormal nonvirulence gene type, in each rice variety of infected rice plants in each rice area under each sampling, the growth index of the strain corresponding to the abnormal nonvirulence gene type is determined according to the growth rate of the number of infected rice plants with the abnormal nonvirulence gene type and the duration of the presence of the abnormal nonvirulence gene type under each sampling. After calculating the difference between the ratio of lesion area of ​​infected rice plants with the abnormal non-virulent gene type in each sampling and the ratio of lesion area of ​​infected rice plants when the abnormal non-virulent gene type first appeared, the mean of all differences in lesion area ratio with the abnormal non-virulent gene type was normalized and used as the lesion area growth rate of the strain corresponding to the abnormal non-virulent gene type in each sampling. When the abnormal non-virulent gene type first appears, the number of different growth stages in time is negatively correlated to obtain the stage propagation influence of the strain corresponding to the abnormal non-virulent gene type. The product of the growth index, lesion area growth rate, and stage spread impact of the strain corresponding to the abnormal non-virulent gene type was calculated to obtain the expansion index of the strain corresponding to the abnormal non-virulent gene on each infected rice plant of each variety in each sampling.

[0006] Furthermore, the method for obtaining the growth index includes: The difference between the number of infected rice plants when the abnormal nonvirulent gene type first appeared and the number of infected rice plants with the abnormal nonvirulent gene type in each sampling was calculated and normalized to obtain the infection scalar number of the strain corresponding to the abnormal nonvirulent gene type in each sampling. The time between each sampling time and the first appearance of the abnormal non-toxic gene type is normalized and used as the propagation time. The ratio of the infection increase to the transmission duration of the abnormal non-virulent gene type in each sampling is used as the growth index of the strain corresponding to the abnormal non-virulent gene type.

[0007] Furthermore, the method for obtaining the growth-affecting indicators includes: For any given strain, obtain the expansion index of the strain corresponding to the abnormal avirulence gene type that the strain infects each rice variety in historical data of each rice-growing area. In each rice-growing area under the current sampling, the ratio between the expansion index of each abnormal non-virulent gene type corresponding to the strain and the mean expansion index of the corresponding abnormal non-virulent gene type in historical data is used as the propagation trend degree of each abnormal non-virulent gene type corresponding to the strain. The product of the negative correlation mapping of the range of the spread trend of all anomalous nonvirulent gene types to the strain and the mean of the spread trend of all anomalous nonvirulent gene types is taken as the historical influence of the strain. The frequency of occurrence of all abnormal avirulent gene types of this strain in historical data was statistically analyzed to determine the infection commonness of this strain. The product of the strain's historical influence and its infection frequency was used as an indicator of the strain's growth impact.

[0008] Furthermore, the method for obtaining the hazard capability index includes: Acquire multi-dimensional environmental data for each rice-growing area; For any strain in any rice-growing area, after calculating the correlation between the environmental data of this rice-growing area and each other rice-growing area in each dimension over time, the mean of the correlation of all dimensions of environmental data is negatively correlated and mapped as the environmental influence degree between this rice-growing area and each other rice-growing area; the difference between the growth influence index of this strain between this rice-growing area and each other rice-growing area is negatively correlated and mapped to obtain the approximate index of the spread of this strain between this rice-growing area and each other rice-growing area. The product of environmental similarity and influence similarity was normalized and used as an approximate indicator of the spread of this strain in this rice-growing area and in each other rice-growing area. The mean values ​​of the approximate spread indicators of this strain in this rice-growing area and all other rice-growing areas were negatively correlated and mapped to serve as the correlation influence degree. The product of the mean growth influence indicators of this strain in all rice-growing areas and the correlation influence degree was normalized and used as the indicator of the strain's harmfulness.

[0009] Furthermore, the method for obtaining the highly transmissible strain includes: For each rice variety, strains with a hazard index greater than the preset hazard threshold are considered as highly transmissible strains for that rice variety.

[0010] Furthermore, the method for obtaining the propagation chain includes: For any strongly transmissible strain of any rice variety, each other rice variety that is also a strongly transmissible strain of the same strain in other rice varieties is used as the analyzed rice plant. At each growth stage before the current sampling, the correlation between the harmful ability index of the highly transmissible strain between the rice variety and the analyzed rice plants is calculated as the harm similarity between the rice variety and the analyzed rice plants at each growth stage; the mean of all the harm similarities at each growth stage before the current sampling is taken as the harm correlation between the rice variety and the analyzed rice plants with the highly transmissible strain. Under this highly transmissible strain, when the hazard association degree is greater than the preset association threshold, and the analyzed rice plant is a kinship marker with the rice plant of this variety, the analyzed rice plant is linked with the rice plant of this variety; after linking, the hazard association degree of the unlinked rice plant varieties is linked iteratively until there are no rice plants of any variety to link, thus obtaining the transmission chain of this highly transmissible strain.

[0011] Furthermore, the method for obtaining the hazard durability includes: For any given strain, the number of strains that spread strongly across all rice varieties is used as the strong spread index of the strain; the number of linked rice varieties in the maximum spread chain corresponding to the strain is used as the spread range index of the strain; and the average value of the damage capacity index of linked rice varieties in the maximum spread chain corresponding to the strain is used as the spread damage degree of the strain. The product of the strain's strong transmissibility index, transmissibility range index, and transmissibility hazard level is taken as the strain's hazard durability.

[0012] Furthermore, adjusting the attention weights of the current strain in the graph neural network based on hazard durability includes: For any strain, the product of the preset initial weight value and the hazard durability after normalization is used as the weight increment of the strain; the sum of the preset initial weight value and the weight increment is used as the attention weight of the strain.

[0013] Furthermore, the abnormal non-toxic gene type refers to the strain species in the infected rice plant and all abnormal non-toxic genes when infecting a certain plant.

[0014] The present invention has the following beneficial effects: This invention analyzes different rice varieties in different rice-growing regions. It analyzes the spread and infection capacity by examining the infection status of strains containing different abnormal non-virulent gene types. Combining historical data, it analyzes the extent of disease spread in the respective rice-growing regions of all strains corresponding to abnormal non-virulent genes and their growth adaptability in different regions to determine the strain's harmful capacity. By comprehensively considering the spread capacity of strains with different gene abnormalities, a more accurate strain harmful capacity is obtained. Based on the phylogenetic associations of different strains with corresponding rice plants, it analyzes the strain's damage durability, considering the spread of phylogenetic infections, and quantifies the long-term stable harmful capacity of strains across phylogenetic varieties. Furthermore, it adjusts the attention weights of the nodes corresponding to each strain based on the obtained damage durability, resulting in a graph neural network, thereby improving the prediction accuracy of effective resistance genes corresponding to rice blast pathogens. This invention analyzes the degree of damage to rice by each strain in rice plants infected with rice blast in different rice-growing regions, and predicts the effective resistance of different rice varieties to this strain based on its stable spread capacity, improving the prediction accuracy of resistance gene effectiveness. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flowchart of an effective disease resistance gene prediction method based on different rice blast fungal communities in different rice-growing areas, provided as an embodiment of the present invention; Figure 2 This is a flowchart illustrating a method for obtaining an expansion index according to an embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of an effective disease resistance gene prediction method based on different rice blast pathogen groups proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] The following description, in conjunction with the accompanying drawings, details the specific scheme of an effective disease resistance gene prediction method based on different rice blast fungal communities provided by the present invention.

[0020] Please see Figure 1 The diagram illustrates a flowchart of an effective disease resistance gene prediction method based on different rice blast fungal communities provided by an embodiment of the present invention. The method includes the following steps: S1: In different rice-growing areas, obtain the growth stage of each rice variety and the ratio of lesion area of ​​infected rice plants at each sampling time; and determine the abnormal non-virulent gene type of each strain in the infected rice plants at each sampling time.

[0021] In this embodiment of the invention, large-scale automated monitoring of rice-growing areas can be achieved by combining drones with a CNN. At a fixed time each day, such as 1 PM, drones acquire monitoring images of each rice-growing area. These images are then input into the CNN, which outputs the growth stage of each rice plant within the area, the number of rice plants currently infected with rice blast fungus, and simultaneously marks the coordinates of the infected plants and the percentage of their lesions. The overall growth stage of a rice-growing area is the stage corresponding to the most rice plants in that area. Rice growth stages include tillering, jointing, booting, heading, and grain filling.

[0022] Taking three target rice-growing areas as an example, for different varieties of rice plants, such as those carrying the disease resistance gene R, rice plants carrying the disease resistance gene R are also obtained through variety records, and the phylogenetic relationships between rice plants of different varieties are labeled to provide a basis for subsequent graph structure construction. Simultaneously, when a rice-growing area as a whole is initially labeled as a certain growth stage by the CNN, that moment is recorded as the start time of that growth stage cycle. Sampling is performed once after the first identification of rice plants infected with rice blast fungus. Subsequently, sampling is performed at a preset frequency in each growth stage of the rice-growing area as a whole, such as once a week, and sampling is also performed simultaneously when the growth stage changes.

[0023] In this embodiment of the invention, during each monitoring and sampling, the coordinates of the identified rice plants infected with rice blast fungus are output to the staff. The staff then sample the lesions on the infected rice plants, isolate and purify the rice blast fungus strains, and perform gene sequencing on each strain to determine its corresponding avirulence gene profile and the abnormality type of each avirulence gene within it; for example, a deletion or mutation of an avirulence gene at a certain location. It should be noted that the specific monitoring and sampling process can be adjusted by the implementer according to the specific implementation scenario, and no further restrictions are imposed here.

[0024] Since the types of abnormal avirulence genes of different strains vary, their pathogenic effects on rice plants also differ. Therefore, taking a single strain corresponding to a certain abnormal type of avirulence gene in a rice plant infected with rice blast of a certain variety as an example, the strains in the infected rice plant and all the abnormal avirulence genes when infecting a certain plant are considered as one type of abnormal avirulence gene.

[0025] For example, if strain i infects a plant and the corresponding atypical avirulent genes include a deletion-type avirulent gene, then the corresponding atypical avirulent gene type is: strain i - deletion-type avirulent gene. In actual rice plants infected with rice blast fungus, it's also possible for a single plant to possess multiple atypical avirulent genes simultaneously. For instance, if strain i infects a plant and the atypical avirulent genes include a mutant avirulent gene and a deletion-type avirulent gene, then the corresponding atypical avirulent gene type is: strain i - mutant avirulent gene and deletion-type avirulent gene. It's understandable that the spread of the strain involves strains containing the relevant avirulent genes spreading to surrounding rice plants via spores. Infectious strains can further spread from infected rice plants to neighboring plants of the same batch, causing them to become diseased as well. In this case, the infected rice plants contain the same relevant avirulent genes as the strain.

[0026] In this embodiment of the invention, multi-dimensional environmental data is monitored simultaneously in each rice-growing area. Temperature, humidity, and pH sensors are installed at monitoring points evenly selected within each rice-growing area. Environmental data for each dimension is recorded hourly, and the timestamps of each sensor are aligned. The environmental data for each dimension in each rice-growing area at a single moment is the average of the environmental data for that dimension monitored by each monitoring point within that rice-growing area at that moment.

[0027] During the spread of rice blast fungus, different strains exhibit varying avirulence gene profiles. Furthermore, the different avirulence gene abnormalities result in varying pathogenicity in rice. Therefore, this study analyzes the stability of the spread of rice blast fungus across different rice-growing regions, the degree of damage to rice plants, and its temporal changes based on a specific strain containing avirulence gene abnormalities. This analysis further analyzes the damage durability of each strain, facilitating subsequent adjustment of attention weights for corresponding nodes of each strain.

[0028] S2: For each infected rice plant of each variety in each rice-growing area under each sampling, the expansion index of the strain corresponding to each abnormal non-virulent gene type is determined based on the existence duration of each abnormal non-virulent gene type of each strain and the increase in the number of infected rice plants, combined with the increase in the lesion area ratio.

[0029] During the spread of rice blast fungus, if a certain avirulent gene in the avirulent gene spectrum of a certain strain can be recognized by the disease resistance gene R of the rice plant, the corresponding rice plant will initiate an immune response. However, if the avirulent gene corresponding to the disease resistance gene R is deleted or mutated, the avirulent gene cannot be recognized by the disease resistance gene R, which may cause the rice plant to become infected.

[0030] When a certain strain causes disease, it may have a certain ability to spread in rice-growing areas, resulting in multiple diseased rice plants in the same rice-growing area and the disease range gradually expanding. Therefore, based on the increase in the infection range over time, we first analyze the degree of transmission of a certain strain to each variety of rice plants in a single rice-growing area during a single sampling.

[0031] For each sampling, the later an abnormal non-virulent gene type appears in a rice-growing area, the slower the growth rate of the number of infected rice plants, and the slower the growth rate of the proportion of lesions in each diseased rice plant, the less the spread of the strain corresponding to the abnormal non-virulent gene type to this rice variety in the rice-growing area.

[0032] Preferably, in this embodiment of the invention, the method for obtaining the expansion index is described in [reference needed]. Figure 2 The diagram illustrates a flowchart of a method for obtaining an expansion index according to an embodiment of the present invention, the method comprising the following steps: S201: For any abnormal non-virulent gene type, in each rice-growing area, for each infected rice plant of each variety, the growth index of the corresponding strain for each abnormal non-virulent gene type is determined based on the growth rate of the number of infected rice plants with the abnormal non-virulent gene type and the duration of the presence of the abnormal non-virulent gene type in each sampling.

[0033] For each variety of infected rice plants, the greater the increase in the number of infected plants in a short period of time, the stronger the short-term spread ability of the abnormal non-virulent gene corresponding to that strain. Therefore, this strain should be given higher attention in the later disease resistance analysis.

[0034] In this embodiment of the invention, the difference between the number of infected rice plants when the abnormal non-virulent gene type first appears and the number of infected rice plants with the abnormal non-virulent gene type in each sampling is calculated and normalized to obtain the infection increment of the strain corresponding to the abnormal non-virulent gene type in each sampling. The larger the infection increment, the stronger the ability of infection to spread.

[0035] The time between each sampling time and the first appearance of the abnormal non-toxic gene type is normalized and used as the propagation time, reflecting the time after the abnormal non-toxic gene type appears and participates in the propagation.

[0036] The ratio of the number of infections corresponding to the abnormal non-virulent gene type to the transmission time in each sampling is then used as the growth index of the strain corresponding to the abnormal non-virulent gene type, reflecting the growth rate of the number of infections. The larger the growth index, the faster the number of infections increases.

[0037] It should be noted that normalization is a technique well known to those skilled in the art. The choice of normalization can be linear normalization or standard normalization, etc., and the specific normalization method is not limited here.

[0038] S202: Based on the growth rate of lesion area in infected rice plants with this abnormal non-virulent gene type, determine the lesion area growth rate of the corresponding strain with this abnormal non-virulent gene type in each sampling.

[0039] Meanwhile, the development speed of lesions on individual infected rice plants was analyzed. For each variety of infected rice plant, when the same abnormal non-virulent gene type exists, the degree of increase in the ratio of lesion area reflects the spread and harm of the strain corresponding to the abnormal non-virulent gene type to this variety of rice plant.

[0040] In this embodiment of the invention, after calculating the difference between the ratio of lesion area of ​​infected rice plants with the abnormal non-virulent gene type in each sampling and the ratio of lesion area of ​​infected rice plants when the abnormal non-virulent gene type first appeared, the mean of the differences in the ratio of lesion area of ​​all infected rice plants with the abnormal non-virulent gene type is normalized and used as the lesion area growth rate of the strain corresponding to the abnormal non-virulent gene type in each sampling. The lesion area growth rate is used to measure the spread and infection capacity of the strain corresponding to the abnormal non-virulent gene in infected rice plants of a single variety.

[0041] It should be noted that the infected rice plants with abnormal non-virulent gene types were all from the same variety of infected rice plants during the analysis, such as rice infected with rice blast fungus carrying the R gene.

[0042] S203: Based on the growth index and the growth rate of lesion area, combined with the growth stage when the abnormal non-virulent gene type first appeared, the expansion index of the strain corresponding to the abnormal non-virulent gene on each infected rice plant of each variety was determined in each sampling.

[0043] Finally, the severity of the short-term spread of the strain corresponding to the abnormal non-virulent gene type to each rice variety in each rice-growing area was quantified from three dimensions: the increase in the number of infections, the development of lesions, and the time of transmissibility.

[0044] In this embodiment of the invention, when the abnormal non-toxic gene type appears for the first time, the number of different growth stages in the time sequence is negatively correlated to obtain the stage propagation influence of the strain corresponding to the abnormal non-toxic gene type. When the abnormal non-toxic gene type appears for the first time, the rice area has already experienced a lot of growth stages in the time sequence, indicating that the degree of propagation influence of the strain corresponding to the abnormal non-toxic gene type is low.

[0045] It should be noted that negative correlation mapping is a technique well known to those skilled in the art. Negative correlation mapping can take the form of inverse proportional value or negative exponent form, etc., which will not be limited or elaborated here.

[0046] Finally, the product of the growth index, lesion area growth rate, and stage spread impact of the strain corresponding to the abnormal non-virulent gene type was calculated to obtain the expansion index of the strain corresponding to the abnormal non-virulent gene on each infected rice plant of each variety in each sampling. The larger the expansion index, the higher the degree of influence of the strain corresponding to the abnormal non-virulent gene on the spread of the single rice plant.

[0047] S3: For each rice variety in each rice-growing area under the current sampling, based on the deviation of the expansion index of each strain corresponding to all abnormal non-toxic gene types from the historical situation, the growth impact index of each strain on each rice variety in each rice-growing area is obtained; based on the similarity of environmental data between each rice-growing area and other rice-growing areas, the deviation distribution of each strain on each rice variety in the growth impact index is analyzed, and the hazard capacity index of each strain is determined.

[0048] Since avirulence gene abnormalities do not necessarily lead to rice disease, for example, a mutation in avirulence gene may give it pathogenicity, but if its adaptability is poor, the strain may not spread effectively when infecting plants, making it less likely to cause large-scale rice blast infection. Furthermore, different strains vary in their harmfulness to rice plants under different environments. Therefore, analyzing historical data on the spread of individual strains in different rice-growing areas is crucial to determining the harmfulness of each strain.

[0049] Among them, the avirulence gene spectrum of a single strain may contain multiple abnormal avirulence genes, and the abnormality types of each abnormal avirulence gene are also different. In the actual propagation process, the expansion and damage capabilities of strains corresponding to each abnormal avirulence gene may be different, and there is a possibility that two abnormalities may have a synergistic effect, which increases their ability to damage rice plants.

[0050] Taking the transmission of a single strain to a single rice variety as an example, if the strain corresponding to all types of abnormal non-toxic genes does not have a serious expansion in a certain rice-growing area, and the historical abnormal frequency in that rice-growing area does not increase significantly, and its diffusion ability is relatively consistent in rice-growing areas under different environmental conditions, then the strain can maintain a low diffusion in multiple environments, and its harmful ability is smaller.

[0051] First, we analyzed the deviation of all strains corresponding to different abnormal non-virulent gene types within a single rice-growing area compared to historical data to measure the degree of influence of the strains on rice plant growth under the current sampling. If, under the current sampling, the spread and diffusion capacity of all strains corresponding to the abnormal non-virulent gene types in the rice-growing area shows a high and relatively consistent increase throughout the historical growth cycle, and the higher the frequency of occurrence of all corresponding abnormal non-virulent gene types in the historical data, the higher the likelihood that the strain is pathogenic to this rice variety, and the greater the impact on rice plant growth.

[0052] Preferably, in this embodiment of the invention, the method for obtaining growth-affecting indicators includes: For any given strain, the expansion index of the strain is obtained by acquiring the abnormal avirulence gene types that the strain infects each variety of rice plants in each rice-growing region in historical data. In this embodiment of the invention, the historical data is the three-year period of the strain infecting single-variety rice plants in each rice-growing region. The specific settings can be adjusted by the implementer according to the specific implementation scenario.

[0053] Furthermore, in each rice-growing area under the current sampling, the ratio between the expansion index of each abnormal non-virulent gene type corresponding to the strain and the average expansion index of the corresponding abnormal non-virulent gene type in historical data is used as the transmission trend degree of each abnormal non-virulent gene type corresponding to the strain. The deviation between the expansion index of the current sampling and the historical average reflects the potential for increased expansion and transmission capacity of the strain under the current sampling. The higher the transmission trend degree, the more likely the strains corresponding to each abnormal non-virulent gene type have an increasing trend of expansion and transmission impact, the higher the harmfulness of the strains, and the more attention should be paid to their infection status.

[0054] Furthermore, the product of the negative correlation mapping of the range of the spread trend degree of all abnormal non-virulent gene types corresponding to the strain and the mean of the spread trend degree of all abnormal non-virulent gene types corresponding to the strain is used as the historical influence degree of the strain. The smaller the range of the spread trend degree of all abnormal non-virulent gene types corresponding to the strain, the more consistent the growth of the spread and influence of the strain. The larger the mean of the spread trend degree of all abnormal non-virulent gene types corresponding to the strain, the higher the overall spread and influence capacity. Therefore, the larger the historical influence degree, the stronger the spread and influence capacity of the strain and the greater the harm.

[0055] Further analysis of the frequency of occurrence of all abnormal non-virulent gene types of this strain in historical data was used as the infection commonness of this strain. The more frequent the occurrence, the stronger the infectivity and the higher the harm.

[0056] Finally, the product of the strain's historical impact and infection frequency was used as the growth impact index of the strain, and the extent to which the strain affected the growth of single-variety rice plants in rice-growing areas was measured based on historical analysis.

[0057] Based on the analysis of multiple rice-growing areas, if the strain has a similar degree of influence on single-variety rice plants in each rice-growing area and under similar conditions with other rice-growing areas, then the growth impact index of the strain is more reliable in characterizing its harmful ability and reflects its more realistic and reliable harmfulness.

[0058] Preferably, in this embodiment of the invention, the method for obtaining the hazard capability index includes: For any strain in any rice-growing region, after calculating the correlation between the environmental data of this rice-growing region and each other rice-growing region in each dimension over time, the mean of the correlation of all dimensions of environmental data is negatively correlated and used as the degree of environmental influence between this rice-growing region and each other rice-growing region. In this embodiment of the invention, the Pearson correlation coefficient is used to calculate the correlation between environmental data of each dimension over time. The overall environmental similarity is low, indicating that the environmental influence between different rice-growing regions is greater. If the similarity of growth influence is still high at this time, it indicates that the strain is less affected by the environment. Since the ability of different environments to affect the fluctuation of transmission hazards is low, excessive hazard attention is not required.

[0059] It should be noted that the Pearson correlation coefficient is a well-known technique in the field and will not be used as an index here.

[0060] A negative correlation mapping was performed between the differences in growth impact indicators of this strain and each other rice-growing area to obtain the approximation of the strain's impact between this rice-growing area and each other rice-growing area. The product of environmental approximation and impact approximation was normalized and used as an approximation index of the strain's spread between this rice-growing area and each other rice-growing area. The larger the spread approximation index, the more similar the strain's growth impact capacity under different environments, and the lower the level of concern regarding its harm can be.

[0061] A negative correlation was established between the average spread approximation index of this strain in this rice-growing area and all other rice-growing areas to represent the correlation influence degree, characterizing the overall environmental impact among rice-growing areas. A smaller overall spread approximation index indicates a higher potential for environmental impact and a greater likelihood of fluctuating damage. The product of the average growth impact index of this strain across all rice-growing areas and the correlation influence degree was normalized to represent the strain's harmfulness. A larger growth impact index corresponds to a larger correlation influence degree, indicating a more significant spread and hazard posed by this strain.

[0062] S4: Under the current sampling conditions, identify the highly transmissible strains of each rice variety based on the hazard capacity index; for each highly transmissible strain of each rice variety, determine the transmission chain of each highly transmissible strain based on its phylogenetic relationship with other rice varieties and the similarity of its hazard capacity index in each growth stage cycle; determine the hazard durability of each strain by the distribution of its hazard capacity index on the maximum transmission chain.

[0063] After analyzing the harmfulness of each strain in single rice varieties, it was found that during the spread of rice blast fungus, there may be strains with strong spread and harmfulness in a certain variety's growth cycle, but they cannot be stably inherited. That is, in the next growth cycle, the strain cannot maintain strong harmfulness in rice plants of that variety or in varieties closely related to that variety.

[0064] Therefore, if a strain can be stably inherited in a certain variety and can cause harm to closely related varieties along the kinship network of that rice variety, and the range of harm gradually increases, the higher the durability of the harm to that variety, the more likely it is to cause serious pathogenic damage to the rice plants.

[0065] Therefore, if a strain has a strong ability to spread stably in a certain rice variety, that is, the higher the strain's harmful ability, and the range of harm caused by the strain gradually increases for varieties that are closely related to it in the kinship network, then the strain's harmful durability will be greater.

[0066] First, strains with high spread and diffusion capabilities are marked, i.e., strong spread strains are screened based on hazard index. In this embodiment of the invention, for each rice variety, strains with hazard index greater than a preset hazard threshold are designated as strong spread strains for that rice variety. Strong spread strains have a high stable spread and diffusion capability. The preset hazard threshold can be set to 0.5, and the specific value can be adjusted by the implementer and is not limited here.

[0067] By observing the spread of highly contagious strains among rice varieties, the extent of the strain's influence on the overall distribution of influence among different rice varieties can be determined. Preferably, in this embodiment of the invention, the method for obtaining the transmission chain includes: For any strongly spreading strain of any rice variety, each other rice variety that is also a strongly spreading strain in other rice varieties is used as the analysis rice plant, and similarity analysis is performed on rice varieties with the same stable and high spread ability.

[0068] At each growth stage before the current sampling, the correlation between the harmful ability index of the highly transmissible strain between the rice variety and the analyzed rice plants is calculated as the similarity of the harm between the rice variety and the analyzed rice plants at each growth stage. In this embodiment of the invention, the Pearson correlation coefficient is used to calculate the correlation. The higher the similarity of the harm, the more similar the influence of the highly transmissible strain at a single growth stage.

[0069] Then, the mean of all the aforementioned harm similarities at each growth stage before the current sampling is taken as the harm correlation between the rice variety and the analyzed rice variety in this highly transmissible strain. The greater the harm correlation, the more likely the two rice varieties are to infect and spread each other.

[0070] Therefore, under this highly transmissible strain, when the hazard correlation is greater than the preset correlation threshold, and the analyzed rice plant and the rice plant of this variety are identified as related, the analyzed rice plant and the rice plant of this variety will be linked. The presence of a related marker indicates that the two rice plants are related, thus demonstrating that the highly transmissible strain does indeed have a continuous spread effect. In this embodiment of the invention, the preset correlation threshold is set to 0.6. The specific value can be adjusted by the implementer and is not limited here.

[0071] At this point, the harm correlation of the unlinked rice varieties is analyzed and linked iteratively until there are no more rice varieties to link, thus obtaining the propagation chain of the highly propagating strain. In this embodiment of the invention, the linked rice varieties are linked to each unlinked rice variety in turn. When no more links are possible, a complete propagation chain is formed. At the same time, all unlinked varieties can be further analyzed to obtain the propagation chain, and finally, several propagation chains of the highly propagating strain can be obtained.

[0072] Analyzing the damage under the maximum transmission chain, the longer the maximum transmission chain and the higher the distribution of the damage index of each variety on the transmission chain, the greater the damage durability.

[0073] Preferably, in this embodiment of the invention, the method for obtaining the hazard durability includes: For any given strain, the number of highly transmissible strains among all rice varieties is used as an indicator of its transmissibility, quantifying the overall transmissibility stability of the strain. The number of linked rice varieties in the maximum transmissibility chain corresponding to the strain is used as an indicator of its transmissibility range, quantifying its transmissibility influence. The average of the damage indicators of the linked rice varieties in the maximum transmissibility chain corresponding to the strain is used as the degree of transmissibility damage of the strain, quantifying its overall transmissibility.

[0074] Finally, the product of the strain's strong transmission index, transmission range index, and transmission hazard degree is taken as the strain's hazard durability. The greater the hazard durability, the more likely the strain is to spread across related species over a long period of time and has a stronger ability to cause damage, thus requiring greater attention to its hazard situation.

[0075] S5: Make predictions by adjusting the attention weights of the current strain in the graph neural network based on hazard durability.

[0076] The final graph structure among rice plants is determined and input into a GNN. The attention weights of the corresponding nodes for each strain are adjusted based on the obtained damage tolerance of each strain. In this embodiment, for any strain, the product of a preset initial weight value and the normalized damage tolerance is used as the weight increment for that strain. The sum of the preset initial weight value and the weight increment is used as the attention weight for that strain, where the initial weight value is set to 1.

[0077] In one specific embodiment of the present invention, the determination of nodes in the graph structure includes: ① taking each strain with a non-toxic gene abnormality as a node, including the location of the abnormal gene, the type of abnormality, and the coordinates of the corresponding rice plant; ② taking each variety of rice infected by rice blast fungus as a node; ③ taking each rice-growing area as a node. If a strain infects a certain variety of rice plant, then the two nodes are connected to establish an edge, and the initial weight of the edge is the proportion of the lesion area of ​​that rice plant. Edges are established between nodes of rice plants with kinship markers, wherein the closer the kinship, the higher the edge weight. For example, the minimum weight of the edge corresponding to a kinship relationship is 1, and the growth step size is 1. It should be noted that the specific training of the graph structure neural network is a well-known technique to those skilled in the art, and will not be described in detail here.

[0078] The adjusted attention weights are assigned to each strain node, and the degree of harm each strain causes to rice carrying a certain resistance gene, such as the resistance gene R, is output. The degree of harm reflects the effectiveness of the resistance. The lower the degree of harm, the higher the effective resistance of the rice plant containing the resistance gene. This results in a more accurate prediction of the effectiveness of the resistance gene. In other words, this prediction system can accurately capture rice plants with durable resistance to rice blast fungus in different rice-growing areas, providing a useful reference for subsequent targeted breeding.

[0079] In summary, this invention analyzes different rice varieties in different rice-growing regions. It analyzes the spread of infection by strains containing different abnormal avirulent gene types, and combines historical data to analyze the extent of disease spread in the respective rice-growing regions and the growth adaptability of each strain in different regions, thus determining its harmfulness. By comprehensively considering the spread of strains with different gene abnormalities, a more accurate assessment of strain harmfulness is obtained. Based on the phylogenetic relationships between different strains and their corresponding rice plants, the invention analyzes the durability of damage caused by each strain, considering the spread of phylogenetic infections, and quantifies the long-term stable harmfulness of strains across phylogenetic varieties. Furthermore, based on the obtained durability of damage, the attention weights of the nodes corresponding to each strain are adjusted to obtain a graph neural network, thereby improving the prediction accuracy of effective resistance genes corresponding to rice blast fungal communities. This invention analyzes the degree of damage caused by each strain to rice in rice plants infected with rice blast in different rice-growing regions, and predicts the effective resistance of different rice varieties to this strain based on its stable spread ability, improving the prediction accuracy of resistance gene effectiveness.

[0080] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0081] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for predicting effective resistance genes based on rice blast pathogen communities in different rice-growing areas, characterized in that, The method includes: In different rice-growing areas, the growth stage of each rice variety and the ratio of lesion area of ​​infected rice plants were obtained at each sampling time; and the abnormal non-virulence gene type of each strain in the infected rice plants at each sampling time was determined. For each infected rice plant of each variety in each rice-growing area in each sampling, the expansion index of each abnormal nonvirulent gene type of each strain was determined based on the existence duration of each abnormal nonvirulent gene type of each strain and the increase in the number of infected rice plants, combined with the increase in the lesion area ratio. For each rice variety in each rice-growing area under the current sampling, the growth impact index of each strain on each rice variety in each rice-growing area is obtained based on the deviation distribution of the expansion index of each strain corresponding to all abnormal non-toxic gene types from the historical situation. Based on the similarity of environmental data between each rice-growing area and other rice-growing areas, the deviation distribution of each strain on each rice variety in the growth impact index is analyzed to determine the hazard capacity index of each strain. Under the current sampling conditions, the highly transmissible strains of each rice variety are identified based on the hazard capacity index. For each highly transmissible strain of each rice variety, the transmission chain of each highly transmissible strain is determined based on its phylogenetic relationship with other rice varieties and the similarity of its hazard capacity index in each growth stage cycle. The hazard tolerance of each strain is determined by the distribution of its hazard capacity index on the maximum transmission chain. Prediction is made after adjusting the attention weights of the current strain in the graph neural network based on hazard durability; The methods for obtaining the growth-affecting indicators include: For any given strain, obtain the expansion index of the strain corresponding to the abnormal avirulence gene type that the strain infects each rice variety in historical data of each rice-growing area. In each rice-growing area under the current sampling, the ratio between the expansion index of each abnormal non-virulent gene type corresponding to the strain and the mean expansion index of the corresponding abnormal non-virulent gene type in historical data is used as the propagation trend degree of each abnormal non-virulent gene type corresponding to the strain. The product of the negative correlation mapping of the range of the spread trend of all anomalous nonvirulent gene types to the strain and the mean of the spread trend of all anomalous nonvirulent gene types is taken as the historical influence of the strain. The frequency of occurrence of all abnormal avirulence gene types of this strain in historical data was counted as the infection commonness of this strain; the product of the historical influence and infection commonness of this strain was used as the growth influence index of this strain. The methods for obtaining the hazard capability index include: Acquire multi-dimensional environmental data for each rice-growing area; For any strain in any rice-growing area, after calculating the correlation between the environmental data of this rice-growing area and each other rice-growing area in each dimension over time, the mean of the correlation of all dimensions of environmental data is negatively correlated and mapped as the environmental influence degree between this rice-growing area and each other rice-growing area; the difference between the growth influence index of this strain between this rice-growing area and each other rice-growing area is negatively correlated and mapped to obtain the approximate index of the spread of this strain between this rice-growing area and each other rice-growing area. The product of environmental similarity and influence similarity was normalized and used as an approximate indicator of the spread of this strain in this rice-growing area and in each other rice-growing area. The mean values ​​of the approximate spread indicators of this strain in this rice-growing area and all other rice-growing areas were negatively correlated and mapped to serve as the correlation influence degree. The product of the mean growth influence indicators of this strain in all rice-growing areas and the correlation influence degree was normalized and used as the indicator of the strain's harmfulness.

2. The method for predicting effective disease resistance genes based on different rice blast pathogen groups according to claim 1, characterized in that, The methods for obtaining the expansion indicators include: For any abnormal nonvirulence gene type, in each rice variety of infected rice plants in each rice area under each sampling, the growth index of the strain corresponding to the abnormal nonvirulence gene type is determined according to the growth rate of the number of infected rice plants with the abnormal nonvirulence gene type and the duration of the presence of the abnormal nonvirulence gene type under each sampling. After calculating the difference between the ratio of lesion area of ​​infected rice plants with the abnormal non-virulent gene type in each sampling and the ratio of lesion area of ​​infected rice plants when the abnormal non-virulent gene type first appeared, the mean of all differences in lesion area ratio with the abnormal non-virulent gene type was normalized and used as the lesion area growth rate of the strain corresponding to the abnormal non-virulent gene type in each sampling. When the abnormal non-virulent gene type first appears, the number of different growth stages in time is negatively correlated to obtain the stage propagation influence of the strain corresponding to the abnormal non-virulent gene type. The product of the growth index, lesion area growth rate, and stage spread impact of the strain corresponding to the abnormal non-virulent gene type was calculated to obtain the expansion index of the strain corresponding to the abnormal non-virulent gene on each infected rice plant of each variety in each sampling.

3. The method for predicting effective disease resistance genes based on different rice blast fungal communities according to claim 2, characterized in that, The methods for obtaining the growth index include: The difference between the number of infected rice plants when the abnormal nonvirulent gene type first appeared and the number of infected rice plants with the abnormal nonvirulent gene type in each sampling was calculated and normalized to obtain the infection scalar number of the strain corresponding to the abnormal nonvirulent gene type in each sampling. The time between each sampling time and the first appearance of the abnormal non-toxic gene type is normalized and used as the propagation time. The ratio of the infection increase to the transmission duration corresponding to the abnormal non-virulent gene type in each sampling is used as the growth index of the strain corresponding to the abnormal non-virulent gene type.

4. The method for predicting effective disease resistance genes based on different rice blast pathogen groups according to claim 1, characterized in that, The method for obtaining the highly transmissible strain includes: For each rice variety, strains with a hazard index greater than the preset hazard threshold are considered as highly transmissible strains for that rice variety.

5. The method for predicting effective disease resistance genes based on different rice blast pathogen groups according to claim 1, characterized in that, The method for obtaining the propagation chain includes: For any strongly transmissible strain of any rice variety, each other rice variety that is also a strongly transmissible strain of the same strain in other rice varieties is used as the analyzed rice plant. At each growth stage before the current sampling, the correlation between the harmful ability index of the highly transmissible strain between the rice variety and the analyzed rice plants is calculated as the harm similarity between the rice variety and the analyzed rice plants at each growth stage; the mean of all the harm similarities at each growth stage before the current sampling is taken as the harm correlation between the rice variety and the analyzed rice plants with the highly transmissible strain. Under this highly transmissible strain, when the hazard association degree is greater than the preset association threshold, and the analyzed rice plant is a kinship marker with the rice plant of this variety, the analyzed rice plant is linked with the rice plant of this variety; after linking, the hazard association degree of the unlinked rice plant varieties is linked iteratively until there are no rice plants of any variety to link, thus obtaining the transmission chain of this highly transmissible strain.

6. The method for predicting effective disease resistance genes based on different rice blast pathogen groups according to claim 1, characterized in that, The method for obtaining the hazard durability includes: For any given strain, the number of strains that spread strongly across all rice varieties is used as the strong spread index of the strain; the number of linked rice varieties in the maximum spread chain corresponding to the strain is used as the spread range index of the strain; and the average value of the damage capacity index of linked rice varieties in the maximum spread chain corresponding to the strain is used as the spread damage degree of the strain. The product of the strain's strong transmissibility index, transmissibility range index, and transmissibility hazard level is taken as the strain's hazard durability.

7. The method for predicting effective disease resistance genes based on different rice blast fungal communities according to claim 1, characterized in that, The adjustment of the attention weights of the current strain in the graph neural network based on hazard durability includes: For any strain, the product of the preset initial weight value and the hazard durability after normalization is used as the weight increment of the strain; the sum of the preset initial weight value and the weight increment is used as the attention weight of the strain.

8. The method for predicting effective disease resistance genes based on different rice blast pathogen groups according to claim 1, characterized in that, The abnormal non-toxic gene types refer to the strains in the infected rice plants and all abnormal non-toxic genes that infect a particular plant.

Images