A method for evaluating crop adaptability based on genotype-environment interaction effect and application thereof

Abstract

This invention discloses a crop adaptability assessment method based on genotype-environment interaction effects and its application, involving the intersection of bioinformatics and artificial intelligence technologies. The model collects phenotypic data of target crops under multiple environments, performs data preprocessing, calculates and centers environmental means, constructs a linear regression model, obtains parameters characterizing adaptability, validates and outputs the model, and classifies adaptability. The intercept measures the average performance of the material, while the slope reflects its responsiveness to environmental changes; both together determine the material's environmental adaptability characteristics. This invention provides a robust method for handling noise and missing values ​​in multi-environment experiments, establishes a unified standard for quantifying adaptability to classify the adaptability of different materials, and helps improve the directionality and accuracy of crop breeding, accelerating the breeding process of high-yield and stable-yield maize varieties.

Description

Technical Field

[0001] This invention relates to the intersection of bioinformatics and artificial intelligence technologies, and in particular to a method for assessing crop adaptability based on genotype-environment interaction effects and its application. Background Technology

[0002] In maize breeding, environmental adaptability is a crucial indicator for assessing the stability and commercial value of maize varieties. Different genotypes exhibit significant differences in growth performance under varying ecological environments. For instance, in hot and arid conditions, some maize genotypes may experience poor growth and drastically reduced yields, while thriving in fertile soil and abundant rainfall. Conversely, other genotypes may maintain relatively stable growth under specific adverse conditions, but perform poorly in other environments. Therefore, accurately assessing the environmental adaptability of maize varieties plays a decisive role in breeding high-yielding and stable-yielding varieties, directly impacting the stability and economic benefits of agricultural production.

[0003] Traditional maize breeding methods have several drawbacks. Firstly, they rely heavily on the breeder's experience and judgment. Breeders subjectively assess the adaptability of maize varieties based on their accumulated practical experience; however, this assessment method is limited by personal knowledge and experience, lacks objective quantitative standards, and makes it difficult to guarantee the accuracy and consistency of the results. Secondly, they primarily rely on data from single environmental trials. Single environmental trials cannot comprehensively simulate maize growth under different ecological environments, ignoring the complex interactions between genotype and environment that affect crop phenotype. This can lead to maize varieties exhibiting poor adaptability when introduced to different environments, failing to meet expected yield and quality standards, and posing potential risks to agricultural production.

[0004] With the development of crop breeding research, multi-environment trials (METs) have become increasingly widely used. METs aim to comprehensively assess the adaptability and stability of crop varieties by testing them under multiple different environmental conditions. However, this method faces many challenges in practical application: significant differences exist in climate, soil, and management practices across different environments, making MET data prone to missing information or noise, thus affecting the reliability of the results. For example, data may be missing in some areas due to sudden natural disasters, or data interference may occur due to inconsistent management practices in different regions, severely impacting the reliability of the results. Furthermore, the evaluation of environmental adaptability lacks unified and clear quantitative standards. Different research or breeding projects employ different evaluation methods and indicators, making effective comparison and integration of research results difficult, limiting the exchange and sharing of breeding experience, and hindering the rapid development of breeding technologies.

[0005] Current breeding practices commonly use methods to screen for high-yielding or highly stable materials, primarily including ranking methods based on phenotypic means under multiple environments and ranking materials individually under a single environment. These methods are simple to operate, but they typically focus only on average yield levels or local performance, ignoring the differences in genotype responses to environmental changes and failing to comprehensively reflect the environmental adaptability characteristics of the materials. In academic research, methods such as the AMMI model (Additive Main Effects and Multiplicative Interaction) and GGE biplot analysis (Genotype and Genotype × Environment interaction) have attempted to analyze the interaction effects between genotype and environment, revealing the stability and specific advantages of materials. However, these methods usually involve complex statistical models with weak parameter interpretation, and the interpretation of results relies on specialized knowledge, making them difficult to widely apply in practical breeding. These limitations make existing analytical methods difficult to accurately, efficiently, and widely use for assessing the environmental adaptability of maize varieties, failing to meet the requirements of precision and efficiency in modern maize breeding.

[0006] In conclusion, establishing a crop adaptability assessment model based on genotype-environment interaction effects is urgently needed. The assessment model proposed in this invention, based on a concise linear regression framework, measures the average performance of materials through the intercept and quantifies their responsiveness to environmental changes through the slope, taking into account both stability and plasticity dimensions. This effectively overcomes the shortcomings of traditional breeding methods and existing analytical methods. Furthermore, it can classify materials into different adaptability categories by analyzing their response characteristics under environmental gradients, thus facilitating the screening of highly adaptable and stable maize varieties. Simultaneously, by combining information from different heterotic groups and optimizing paternal selection and hybridization strategies, the performance of target traits can be further improved, which has significant practical implications for enhancing the efficiency and accuracy of maize breeding. Summary of the Invention

[0007] The purpose of this invention is to provide a crop adaptability assessment method based on genotype-environment interaction effects and its application, to provide technical support for the precise breeding of adaptable varieties and to help cultivate high-yield and high-quality maize.

[0008] To solve the above-mentioned technical problems, the invention adopts the following technical solution.

[0009] An efficient crop adaptability assessment method based on genotype-environment interaction effects includes the following steps:

[0010] a. Data acquisition and preprocessing: Collect various phenotypic data of the target crop under multiple environments, remove outliers from the phenotypic data, and perform numerical transformation on the original phenotypic data to make the distribution closer to a normal distribution;

[0011] b. Environmental mean calculation and centering: Calculate the phenotypic mean of all materials under each environment, and use it as a quantitative indicator of the main environmental effect in the evaluation model; center the environmental mean to eliminate systematic bias between environments;

[0012] c. Linear Regression Model Construction: A linear regression model is constructed for each material separately, with the centered environmental mean as the independent variable and the phenotypic value of the material in each environment as the dependent variable. An algorithm is constructed to fit the model. The specific algorithm is as follows: the phenotypic value of a material in a certain environment is equal to the phenotypic value of the material when the environmental mean is zero, i.e., the intercept, plus the product of the material's response strength to environmental changes (i.e., the slope) and the environmental mean, and finally, a residual is added.

[0013] d. Obtain parameters characterizing adaptability: quantify the average performance of the material by the intercept; the larger the intercept value, the higher the phenotypic level of the material under overall environmental conditions; quantify the plasticity of the material by the slope; the larger the absolute value of the slope, the more sensitive the material is to environmental changes, i.e., the higher the plasticity; plot a two-dimensional scatter plot with the intercept and slope as coordinates, and divide the material into four adaptive regions based on the intercept and slope;

[0014] e. Model Validation and Output: The goodness of fit of the model is evaluated through residual analysis and the coefficient of determination R² to ensure the accuracy and stability of the evaluation results; key parameters are output, including intercept, slope, residual and other information for each material, and corresponding visualization charts are generated to intuitively show the performance trends of different materials in multiple environments;

[0015] f. Adaptability Classification: A two-dimensional coordinate system is constructed based on two parameters, intercept and slope, to classify materials into four categories: broad adaptability, stable adaptability, special adaptability and low adaptability, and generate a visual chart.

[0016] As a preferred embodiment of the present invention, the numerical conversion in step a specifically comprises:

[0017] Take the absolute value of all phenotypic values, then add 1 to the absolute value, and then perform a logarithmic transformation to base 10 to linearize the data distribution.

[0018] As a preferred embodiment of the present invention, the method for dividing the region of the two-dimensional scatter plot in step d includes:

[0019] The intercept and slope parameters of the material were obtained based on a linear regression model and then standardized. The intercept was standardized using Z-score to ensure its mean was 0; the slope was standardized using a 1-centered method to measure its stability against deviations from 1, thus unifying the reference standard for different properties. Within an environmental range, the overall mean of the intercept was used as the vertical axis boundary, and the mean of the slope as the horizontal axis boundary. Preliminary classification was performed based on the intersection of the fitted line and the mean fitted line (i.e., the slope of 1).

[0020] 1) For materials that do not intersect, divide them into region ① and region ② with the mean intercept value of 0 as the dividing line;

[0021] 2) For materials that intersect, divide them into region ③ and region ④ with the mean slope of 0 as the dividing line;

[0022] The materials in each region exhibit specific patterns. Among them, the materials in region ② show broad adaptability and possess the target breeding traits. Therefore, when drawing a two-dimensional scatter plot based on yield traits, for traits that are negatively correlated with yield, the intercept is reversed to ensure that the target traits are classified into region ②.

[0023] As a preferred embodiment of the present invention, the four types of adaptive regions in step d are specifically as follows:

[0024] Region ①: Intercept below 0, i.e., when the intercept is below the average performance, it indicates low fitness and poor adaptability to all environments;

[0025] Region ②: Intercept above 0, i.e., when the intercept is higher than the average performance, it represents broad adaptability and is suitable for all environments;

[0026] Region ③: The slope is significantly greater than 0, that is, when the slope is significantly greater than the average plasticity, it represents environmental sensitivity and adaptation to favorable conditions;

[0027] Region ④: The slope is significantly less than 0, that is, when the slope is less than the average plasticity, it represents the stability of adversity, and the potential cannot be realized under favorable conditions;

[0028] The average performance is the mean of the intercepts of all materials, and the average plasticity is the mean of the slopes of all materials.

[0029] A quantitative analysis method for crop environmental adaptability based on the interaction effect between genotype and environment, wherein the quantitative analysis of crop environmental adaptability is performed based on any of the assessment methods described above.

[0030] As a preferred technical solution of the present invention, the following steps are included:

[0031] Genotype-environment interaction effect decomposition: Phenotypic values ​​are decomposed into genotype main effect, environmental main effect, and genotype-environment interaction effect. Specifically, the phenotypic value of a crop variety under a specific environment is equal to a pre-defined value representing the overall average level, plus the value of the genotype main effect of that crop variety, plus the value of the environmental main effect, plus the value of the genotype-environment interaction effect, and finally, a residual value reflecting the deviation between the actual measured value and the predicted value calculated through analysis is added. The environmental main effect is captured through the environmental mean, and the genotype-environment interaction effect is linearly fitted using a slope.

[0032] Calculation of environmental adaptability parameters: The intercept and slope are extracted using a linear regression model to characterize the basic phenotypic value and environmental response intensity of the material, respectively; Phenotypic plasticity is defined as follows: Plasticity is equal to the absolute value of the slope, and the higher the index, the stronger the plasticity of the material.

[0033] Adaptive dynamic analysis: Constructing a material adaptability matrix by combining intercept and slope parameters;

[0034] Materials are classified into four categories based on adaptability: broadly adaptable, particularly adaptable, poorly adaptable, and stable adaptable.

[0035] As a preferred technical solution of the present invention, the linear fitting method for the genotype-environment interaction effect includes: constructing an algorithm to optimize model parameters. The specific algorithm is as follows: the basic phenotypic value of a material under standard environment is equal to the phenotypic value of the material when the environmental mean is zero, i.e., the intercept, plus the product of the material's response intensity to the environment (i.e., the slope) and the environmental mean, and finally adding a residual; verifying the significance of the slope through hypothesis testing, and screening for statistically significant environmental adaptation parameters.

[0036] This case also includes a method for breeding high-yielding and high-quality maize varieties based on environmental adaptability analysis, comprising the following steps:

[0037] Material selection strategy: From the four types of adaptability regions, the widely adaptable materials in region ② are given priority as parents; for the particularly adaptable materials in region ③, targeted improvements are made to target their high plasticity.

[0038] Matching materials - environmental analysis: Materials from region ③ are preferred for planting in a superior environment to fully realize their phenotypic potential; materials from region ④ are suitable for adverse environments, and their high stability ensures basic yield.

[0039] Hybrid combination optimization: Under favorable conditions, materials from region ③ are selected as the male parent to maximize hybrid vigor; under adverse conditions, materials from region ④ are selected as the male parent to ensure hybrid stability.

[0040] Utilization of heterotic groups: Selecting specific paternal parents to optimize target traits based on the adaptive characteristics of heterotic groups;

[0041] Phenotypic validation and iterative breeding: Verify the adaptability of hybrids through multi-environment phenotypic testing;

[0042] The parent selection and pairing strategies are iteratively optimized based on the test results.

[0043] As a preferred technical solution of the present invention, the targeted improvement includes: introducing highly evaluable model plasticity traits into low-fit materials in region ① through gene editing or hybridization; and fixing the stress-stability genes of stable-fit materials in region ④ through phenotypic selection.

[0044] The beneficial effects achieved by adopting the above technical solution are as follows: As detailed in the experimental examples and related figures below, this invention provides a crop adaptability assessment method based on genotype-environment interaction effects and its application. It quantifies environmental adaptability by combining linear regression analysis and employs an environmental adaptability grading method, providing an efficient and accurate environmental adaptability assessment tool. This method not only effectively addresses noise and missing data in multi-environment experimental data but also grads the adaptability of materials based on their response characteristics under environmental gradients, significantly enhancing the directionality and accuracy of the breeding process, thereby accelerating the breeding process of high-yield and stable-yield maize varieties and improving breeding efficiency. Attached Figure Description

[0045] Figure 1 This is a flowchart of the breeding process for high-yield and high-quality maize varieties based on environmental adaptability analysis.

[0046] Figure 2 A schematic diagram illustrating the division of environmental adaptation zones at the population level;

[0047] Figure 3 An analysis diagram of the environmental adaptability of a single material with multiple properties;

[0048] Figure 4 This is a diagram illustrating the adaptability analysis of hybrids with the male parent as the research subject.

[0049] Figure 5 Adaptability analysis diagram of hybrids produced by different combinations of paternal parents from heterotic groups;

[0050] Figure 6 A schematic diagram of the input data format for phenotypic data preprocessing;

[0051] Figure 7 A diagram illustrating the input data format for obtaining intercept and slope data;

[0052] Figure 8 A schematic diagram of the data frame containing the intercept and slope;

[0053] Figure 9A schematic diagram of the input data format for determining the correlation between other traits and ear weight trait;

[0054] Figure 10 A schematic diagram showing the results of the environmental adaptability classification of traits;

[0055] Figure 11 This is a schematic diagram showing the environmental adaptability classification results of a single material with multiple properties. Detailed Implementation

[0056] The following embodiments illustrate the present invention in detail. All raw materials and equipment used in the present invention are conventional commercially available products and can be directly obtained through market purchase. In the following description of the embodiments, specific details such as particular system structures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that the present application can also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of this application with unnecessary detail.

[0057] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of a described feature, integral, step, operation, element, and / or component, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. It should also be understood that, as used in this specification and the appended claims, the term "and / or" refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0058] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0059] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance. References to "one embodiment" or "some embodiments" in this application mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0060] This invention provides a crop adaptability assessment method based on genotype-environment interaction effects and its application. This model combines phenotypic data modeling with the principles of quantitative genetics to transform the environmental response characteristics of materials into quantifiable parameters, providing technical support for the precise breeding of adaptable varieties and helping to cultivate high-yield and high-quality maize.

[0061] Example 1

[0062] A method for assessing crop adaptability based on genotype-environment interaction effects includes the following steps: Figure 1 As shown:

[0063] (1) Phenotypic data of inbred lines and hybrids from various ecological zones, ranging from spring-sown maize in northern China to summer-sown maize in the Huang-Huai-Hai Plain, were collected. To obtain phenotypic data of inbred lines and hybrids under different environments, planting trials were conducted in five ecological zones: Jilin, Liaoning, Beijing, Hebei, and Henan, in 2014 and 2015, respectively. Phenotypic data of 1,404 maize inbred lines in 2014 and 6,210 hybrids in 2015 were collected under the same five environments. Phenotypic data of inbred lines included traits such as ear weight, 100-kernel weight, kernel length, kernel width, average kernel projection area, test weight, and starch content, while phenotypic data of hybrids included traits such as ear weight, plant height, and flowering time. The phenotypic data frame for inbred lines contained 7020 rows and 32 columns, with the first column representing the material name, the second the environment name, and the remainder the phenotypic data. The phenotypic data frame for hybrids contained 31050 rows and 5 columns, representing the material name, environment name, flowering date, plant height, and ear weight, respectively. The missing data rates were 2.9% and 4.0%, respectively. The missing data rates for both inbred lines and hybrids were below 5%, with negligible impact, indicating high data completeness and effectively improving the reliability of environmental adaptability analysis. Large-scale multi-phenotypic determination helps to identify key genetic loci, providing strong support for genetic analysis, environmental adaptability assessment, and breeding decisions.

[0064] (2) Preprocess the phenotypic data to ensure data quality and standardization. First, observe the distribution of the phenotypic data and filter the phenotypic values ​​of the 30 traits to remove outliers to ensure data consistency. The phenotypic data of crop traits usually conform to a normal distribution, but outliers may appear under extreme environmental conditions, which need to be judged within a reasonable range.

[0065] To ensure data validity and improve the reliability of subsequent analysis, this study employs the SSR (Studentized Residuals) method to filter outliers. This method is suitable for normally distributed data and assesses the anomalies of data points based on the standardization of residuals. For each trait, using multi-environment data, the lmer function from the lme4 package in R is used for linear fitting to calculate the BLUE (Best Linear Unbiased Estimator) value. Material is set as a fixed effect, and environment is set as a random effect. This method is suitable for data analysis of balanced designs and can improve the accuracy of trait assessment. The residuals for each phenotypic value are then calculated based on the fitting results. A significance level of α = 0.005 is set, and the residuals are standardized. Data with standardized residuals exceeding the range [-2.807, 2.807] are removed to eliminate potential outliers. Materials with residuals distributed within the range [-2.807, 2.807] are retained to ensure data validity and the reliability of subsequent analysis. Specific implementation steps include (taking ear weight as an example):

[0066] Input data format as follows Figure 6 As shown.

[0067] #Filter outliers

[0068] sigle_trait1 <- trait_filter30 %>%

[0069] filter(trait == "EW")

[0070] sigle_trait1 <- na.omit(sigle_trait1)

[0071] ml <- lmer(value ~ line + (1 | prov), data = sigle_trait1)

[0072] residuals_ml <- as.data.frame(scale(residuals(ml)))

[0073] residuals_ml$judge <-

[0074] ifelse(residuals_ml$V1 >= -2.807 & residuals_ml$V1 <= 2.807, 1, 0)

[0075] delete_position <- which(residuals_ml$judge == 0)

[0076] sigle_trait_noOutliers2 <- sigle_trait1[-delete_position, ].

[0077] (3) Transform the phenotypic data to meet the requirements of biological and statistical analysis. Calculate the absolute value of all phenotypic values, and then add 1 to the absolute value to ensure the reasonableness of the data and avoid the influence of negative values ​​on subsequent calculations. Subsequently, perform a base-10 logarithmic transformation on the data (log 10 (y+1)), this transformation method can compress the range of data values, making the data closer to a normal distribution, while ensuring that zero-value data can be successfully transformed, avoiding numerical loss or calculation errors. For interphase traits related to phenological periods, since their phenotypic values ​​may be negative, direct logarithmic transformation may lead to calculation errors. Therefore, the absolute value is taken first to ensure the rationality of the data and to make the smaller value reflect more favorable fertilization or flowering synchronization. Subsequently, logarithmic transformation is performed to further improve the normality of the data, improve the model fitting effect, and meet the assumptions of statistical methods such as linear regression. This method helps to reduce the skewness of the data, enhance the stability of the analysis, and improve the prediction accuracy. Specific implementation steps include (taking ear weight trait as an example):

[0078] #Phenological value preprocessing

[0079] sigle_trait_noOutliers2 <- trait_filter28_noOutliers[trait_filter28_noOutliers$trait == "EW",]

[0080] sigle_trait_noOutliers2$value <- abs(sigle_trait_noOutliers2$value)

[0081] sigle_trait_noOutliers2$value <- log10(sigle_trait_noOutliers2$value+1)

[0082] (4) Construct a linear regression model and obtain parameters characterizing environmental adaptability.

[0083] A linear regression model is constructed to obtain intercept and slope data frames, quantifying the average performance and average plasticity of the material. A quantitative analysis method for crop environmental adaptability based on genotype-environment interaction effects (G×E) is proposed, applicable to genetic improvement and adaptive breeding of crops such as maize and rice. The core of the method lies in quantifying the response characteristics of crop materials to the environment through a linear regression model, specifically including the following steps: For phenotypic data (e.g., yield) of the target crop material under multiple environments (e.g., different regions or years), the phenotypic mean for each environment is calculated and centered as a quantitative indicator of environmental effect. ); A separate linear regression model was built for each material, using the centered environmental mean ( The independent variable is , and the phenotypic values ​​of the material under various environments are the dependent variable. An algorithm (such as the lm() function in R) is constructed to fit the model, and the formula is: y ij Let i be the phenotypic value of material i in environment j. Let βj be the environmental mean of environment j, β0i be the intercept, representing the phenotypic value of material i when the environmental mean is zero, β1j be the slope, reflecting the intensity of material i's response to the environment (plasticity), and εij be the residual (including random error and unmodeled G×E nonlinear effects). It is important to note that when fitting the model, the environmental mean should be centered so that the intercept term represents the phenotypic value of the material under average conditions, reducing the influence of differences in environmental mean at different experimental points on the regression coefficients. Plasticity refers to the magnitude of phenotypic change of a genotype in different environments, usually expressed as a slope βj. 1j Quantification. The steeper the slope, the more sensitive the material is to environmental changes (high plasticity); a slope close to zero indicates phenotypic stability (low plasticity). This method combines phenotypic data modeling with the principles of quantitative genetics to transform the environmental response characteristics of materials into quantifiable parameters, providing technical support for the precise breeding of adaptable varieties.

[0084] In classical quantitative genetics, phenotype is typically broken down into: G i E represents the main effect of genotype. j Represents the main environmental effect, (G×E) ij This represents the interaction effect between genotype and environment. In our model, the environmental mean... Implicit environmental main effect E j And the slope β 1j Then (G×E) was partially captured. ij The linear component.

[0085] Using the `lm` function in R, with the environmental mean as the independent variable (X) and the phenotypic value as the dependent variable (Y), a linear fit was performed on each material and trait to obtain the intercept and slope parameters. The intercept reflects the average performance of the material, i.e., the overall level under all environments, while the slope measures the material's response to environmental changes, i.e., its plasticity. Finally, the intercepts and slopes of all materials were compiled into a data frame to provide a basis for subsequent adaptive analysis and regional division. The construction details are as follows:

[0086] Input data format as follows Figure 7 As shown.

[0087] # Obtain intercept and slope data frames

[0088] get_parameters_df <- sigle_trait_noOutliers2[!sigle_trait_noOutliers2$line %in% delet_lines,]

[0089] envMean_df <- get_parameters_df %>%

[0090] group_by(prov) %>%

[0091] summarize(Env_Mean = mean(value, na.rm = TRUE)) %>%

[0092] as.data.frame()

[0093] envMean_df$Env_Mean <- as.numeric(scale(envMean_df$Env_Mean,center =T,scale = F))

[0094] get_parameters_df <- envMean_df %>%

[0095] left_join(get_parameters_df,by = "prov")

[0096] get_parameters_df_allTraits <- rbind(get_parameters_df_allTraits,get_parameters_df)

[0097] fit_params <- get_parameters_df %>%

[0098] group_by(line) %>%

[0099] summarize(Intercept = coef(lm(value ~ Env_Mean ))[1],

[0100] Slope = coef(lm(value ~ Env_Mean))[2]) %>%

[0101] arrange(desc(Intercept)) %>% # Sort by intercept in descending order

[0102] as.data.frame()

[0103] fit_params$trait <- rep(j,dim(fit_params)[1])

[0104] Output data format as follows Figure 8 As shown.

[0105] #Calculate the correlation between other traits and yield traits. The correlation results will be used for subsequent data transformation to draw a two-dimensional scatter plot. Input data format is as follows: Figure 9 As shown.

[0106] GY_mydata <- get_parameters_df_allTraits

[0107] envs <- unique(GY_mydata$prov)

[0108] correlation_df <- data.frame()

[0109] for(i in 1:length(envs)) {

[0110] temp_trait <- GY_mydata[GY_mydata$prov == envs[i], ]

[0111] temp_trait <- temp_trait[,c("line","trait","value")]

[0112] temp_trait <- spread(temp_trait,trait,value)

[0113] df_cor <- temp_trait %>%

[0114] select(-1) %>% # Delete the first column (if needed)

[0115] summarise(across(-"EW", # Calculate correlation for each column except "EW").

[0116] ~ {

[0117] result <- cor.test(.x, temp_trait$EW, use = "complete.obs")

[0118] tibble(cor = result$estimate, p_value = result$p.value)

[0119] })) %>%

[0120] pivot_longer(cols = everything(), names_to = "trait", values_to = "value") %>%

[0121] unnest(value) %>% # Expand nested tibble

[0122] as.data.frame()

[0123] df_cor$env <- rep(envs[i],dim(df_cor)[1])

[0124] df_cor <- df_cor[order(df_cor$cor),]

[0125] correlation_df <- rbind(correlation_df,df_cor)

[0126] }

[0127] correlation_df <- correlation_df[order(correlation_df$p_value),]

[0128] correlation_df_neagtive <- correlation_df[correlation_df$cor < 0,]

[0129] correlation_df_neagtive <- correlation_df_neagtive[correlation_df_neagtive$p_value <= 0.05,]

[0130] (5) Based on two-dimensional scatter plots, materials are adaptively classified using specific standards.

[0131] Materials are adaptively classified using specific criteria based on two-dimensional scatter plots. Intercept and slope parameters are obtained through linear fitting and then standardized. The intercept is standardized using Z-score to a mean of 0; the slope is standardized using a 1-centered method to also have a mean of 0, but here 0 represents a slope of 1, measuring its stability against deviations from 1, thus unifying the reference standard for different properties. Figure 2 As shown, within a specific environmental range, preliminary classification is made based on the intersection of the fitted line and the mean fitted line (mean slope = 1) of the material: 1) For materials without intersection, the mean intercept (0) is used as the dividing line to divide them into region ① and region ②; 2) For materials with intersection, the mean slope (0) is used as the dividing line to divide them into region ③ and region ④.

[0132] The materials from each region exhibit specific patterns. Among them, the materials from region ② show broad adaptability and possess breeding target traits, such as earlier flowering and higher ear weight. Therefore, when plotting scatter plots based on yield traits, for traits negatively correlated with yield, the intercept is reversed to ensure that target traits (such as early flowering materials) are classified into region ②.

[0133] Based on the above criteria, the adaptability of materials can be divided into four regions, each with different biological significance, such as... Figure 2 As shown in the figure, region ① has an intercept below 0 (average performance), indicating low fitness, meaning the material has poor adaptability in all environments; region ② has an intercept above 0 (average performance), representing broad fitness, meaning the material can adapt to various environments; region ③ has a slope significantly greater than 0 (average plasticity), representing environmental sensitivity, meaning the material performs well in favorable conditions but is relatively sensitive to environmental changes; region ④ has a slope significantly less than 0 (average plasticity), representing adversity stability, meaning the material's potential is not fully realized in favorable conditions but it can maintain stable performance in adversity.

[0134] (6) Model Validation and Output

[0135] like Figure 2 As shown, Material 2 has a phenotypic value higher than the average, corresponding to region ② in the adaptability analysis diagram, and is adapted to all environments; Material 3 is sensitive to the environment, corresponding to region ③ in the adaptability analysis diagram, and is adapted to favorable environments; Material 1 has a phenotypic value lower than the average, corresponding to region ① in the adaptability analysis diagram, and has poor adaptability to all environments; Material 4 is not sensitive to the environment, corresponding to region ④ in the adaptability analysis diagram, and is adapted to unfavorable environments, failing to realize its growth potential in favorable environments. In summary, materials located in region ② have general adaptability, while materials located in regions ③ and ④ are adapted to specific environments.

[0136] Adaptability analysis diagrams were drawn to classify the adaptability of materials, and their environmental adaptation types were determined based on the distribution characteristics of intercepts and slopes. Target breeding traits preferentially clustered in region ② (widely adaptable region) to ensure that materials exhibit stable and excellent performance under different environments. To improve the efficiency and operability of adaptability assessment, a unified classification standard was established to quickly identify materials with excellent adaptability, simplify the screening process in breeding practice, and enhance the practical application value of the method. The model construction details are as follows:

[0137] plot_mean_df <- as.data.frame(matrix(nrow=5,ncol=5))

[0138] colnames(plot_mean_df) <- c("prov","Env_Mean","line","trait","value")

[0139] plot_mean_df$prov <- envMean_df_meanNoCentered$prov

[0140] plot_mean_df$Env_Mean <- envMean_df$Env_Mean

[0141] plot_mean_df$line <- rep("average",5)

[0142] plot_mean_df$trait <- rep(j,5)

[0143] plot_mean_df$value <- envMean_df_meanNoCentered$Env_Mean #valueshould be environment mean that not be centered

[0144] average_para <- plot_mean_df %>%

[0145] group_by(line) %>%

[0146] summarize(Intercept = coef(lm(value ~ Env_Mean ))[1],

[0147] Slope = coef(lm(value ~ Env_Mean))[2]) %>%

[0148] arrange(desc(Intercept)) %>% # Sort in descending order by intercept

[0149] as.data.frame()

[0150] average_para$trait <- j

[0151] colorSet <- c('#483ba3','#ffa8a8')

[0152] fit_params <- fit_params[order(fit_params$Intercept),]

[0153] colors <- colorRampPalette(colorSet)(nrow(fit_params))

[0154] plot_get_parameters_df <- get_parameters_df

[0155] plot_get_parameters_df <- rbind(plot_get_parameters_df,plot_mean_df)

[0156] fit_params_intercept_scale <- fit_params

[0157] fit_params_intercept <- rbind(fit_params_intercept_scale,average_para)

[0158] fit_params_intercept_scale$Intercept <- as.numeric(scale(fit_params_intercept_scale$Intercept))

[0159] fit_params_intercept_scale$Slope <- (fit_params_intercept_scale$Slope- 1) / sd(fit_params_intercept_scale$Slope)#3)slope scale

[0160] df_to_judge_crossover <- merge(plot_get_parameters_df,fit_params_intercept,by = c("line","trait"),all.x=T)

[0161] df_to_judge_crossover$cross_flag <- 0

[0162] ref_params <- df_to_judge_crossover[df_to_judge_crossover$line == "average", ]

[0163] a_ref <- unique(ref_params$Slope)

[0164] b_ref <- unique(ref_params$Intercept)

[0165] materials <- unique(df_to_judge_crossover$line)

[0166] materials <- materials[materials != "average"]

[0167] for (mat in materials) {

[0168] mat_params <- df_to_judge_crossover[df_to_judge_crossover$line ==mat, ]

[0169] if (nrow(mat_params) > 0) {

[0170] a_mat <- unique(mat_params$Slope)

[0171] b_mat <- unique(mat_params$Intercept)

[0172] if (a_mat != a_ref) {

[0173] x_star <- (b_ref - b_mat) / (a_mat - a_ref)

[0174] x_range <- df_to_judge_crossover[df_to_judge_crossover$line ==mat, "Env_Mean"]

[0175] if (min(x_range) <= x_star & x_star <= max(x_range)) {

[0176] df_to_judge_crossover$cross_flag[df_to_judge_crossover$line == mat] <- 1

[0177] }

[0178] }else{

[0179] print(paste0(j,"exit slope equal!!"))

[0180] }

[0181] }

[0182] }

[0183] no_corssover <- df_to_judge_crossover[df_to_judge_crossover$cross_flag == 0,]

[0184] ref_intercept <- no_corssover[no_corssover$line == "average",]

[0185] ref_intercept <- ref_intercept$Intercept[1]

[0186] if(j %in% c( "DTS", "SAI", "STI", "DTA", "DTT", "LBT", "ATI")){

[0187] zone2 <- no_corssover[no_corssover$Intercept <= ref_intercept,]

[0188] } else{

[0189] zone2 <- no_corssover[no_corssover$Intercept >= ref_intercept,]

[0190] }

[0191] zone2_lines <- unique(zone2$line)

[0192] zone2_lines <- zone2_lines[zone2_lines != "average"]

[0193] if(j %in% c( "DTS", "SAI", "STI", "DTA", "DTT", "LBT", "ATI")){

[0194] zone1 <- no_corssover[no_corssover$Intercept > ref_intercept,]

[0195] } else{

[0196] zone1 <- no_corssover[no_corssover$Intercept < ref_intercept,]

[0197] }

[0198] zone1_lines <- unique(zone1$line)

[0199] have_corssover <- df_to_judge_crossover[df_to_judge_crossover$cross_flag == 1,]

[0200] zone4 <- have_corssover[have_corssover$Slope < 1,]

[0201] zone4_lines <- unique(zone4$line)

[0202] zone3 <- have_corssover[have_corssover$Slope > 1,]

[0203] zone3_lines <- unique(zone3$line)

[0204] line_list <- list(zone1_lines,zone2_lines,zone3_lines,zone4_lines)

[0205] names(line_list) <- c("zone1","zone2","zone3","zone4")

[0206] fit_params_intercept_scale_color <- fit_params_intercept_scale

[0207] fit_params_intercept_scale_color <- fit_params_intercept_scale_color%>%

[0208] mutate(Color = case_when(

[0209] line %in% line_list$zone1 ~ "1",

[0210] line %in% line_list$zone2 ~ "2",

[0211] line %in% line_list$zone3 ~ "3",

[0212] line %in% line_list$zone4 ~ "4",

[0213] TRUE ~ "black" ))

[0215] if(j %in% c( "DTS", "SAI", "STI", "DTA", "DTT", "LBT", "ATI")){

[0216] fit_params_intercept_scale_color$Intercept <- -fit_params_intercept_scale_color$Intercept

[0217] } else{

[0218] }

[0219] Model validation and output results are as follows Figure 10 As shown.

[0220] like Figure 3 As shown, using materials as the research object, a two-dimensional scatter plot of adaptability for multiple traits is drawn. First, the intercept and slope of each material on all traits are calculated using the following code, and a data frame is generated. Then, the ggplot function in the ggplot2 package of R language is used to visualize the data and draw scatter plots to intuitively show the environmental adaptability characteristics of each material on different traits.

[0221] load("E: / work_report / 20250223 / data_analysis / triangle / Intercept_slope_scale.rdata")#slope scale

[0222] Intercept_change <- fit_params_intercept_scale_color_allTraits

[0223] Intercept_change <- Intercept_change %>%

[0224] mutate(Intercept = if_else(grepl("DTS|SAI|STI|DTA|DTT|LBT|ATI",trait), -Intercept, Intercept))

[0225] temp_traitDefinition

[0226] line_traits <- as.data.frame(table(Intercept_change$line))#1398

[0227] line_25traits <- line_traits[line_traits$Freq >= 25,]#1297

[0228] line_25traits <- line_25traits$Var1

[0229] tempPlot <- Intercept_change[Intercept_change$line %in% line_25traits,]

[0230] The data frame (tempPlot) for the environmental adaptability analysis of multiple properties of a single material is as follows: Figure 11 As shown.

[0231] Example 2

[0232] Based on the above research methods, an adaptive analysis was conducted on the phenotypic data of 30 traits of inbred lines and 3 traits of hybrids under 5 environments.

[0233] On the one hand, we focus on the materials themselves, retaining those with 25 or more traits. This is because not all materials have complete phenotypic data for all traits, and retaining more traits helps improve the reliability of the analysis and ensures the comprehensiveness of the fitness assessment. For each material, all traits are divided into four categories, and improvements are made to the traits of different fitness categories to enhance the overall performance of the material. For example... Figure 3 Taking two materials as examples, for material MG_1000, 18 traits are located in region ②. Materials with broad adaptability in region ② are preferred as parents, as most traits are broadly adaptable. Therefore, traits located in other regions can be improved to enhance the material's adaptability and overall performance. Material MG_1422, on the other hand, has 13 traits located in region ③, exhibiting specific adaptability and superior performance under favorable conditions. Targeted improvement can be carried out on its highly plastic traits. For low-fitness materials in region ①, high-plasticity traits can be introduced through gene editing or hybridization. For stable-fitness materials in region ④, adversity-stability genes can be fixed through phenotypic selection.

[0234] On the other hand, considering traits, different materials should be selected for different environments. Based on the above findings, we developed a method for breeding high-yielding and high-quality maize varieties based on environmental adaptability analysis. For example, for favorable environments, we should select materials located in region ③, which are sensitive to the environment and can reach their maximum potential; for unfavorable environments, we should select materials located in region ④, which are not sensitive to the environment, have good stability, and ensure their basic survival (e.g., ...). Figure 2 At the same time, we can also adjust the mating methods to breed hybrids for different environments (e.g., Figure 4 In favorable environments, materials from region ③ that are sensitive to environmental conditions can be selected as the male parent to optimize the hybrid's phenotype. In unfavorable environments, materials from region ④ that are not sensitive to environmental conditions can be selected as the male parent to improve the stability of the hybrid. KWS49's F1 generation flowers early and is adaptable to all environments; using KWS49 as the male parent for hybridization with the female parent can yield a hybrid with good adaptability and early flowering. 5831's F1 generation is short; using 5831 as the male parent for hybridization with the female parent population can yield a dwarf hybrid. DH351's F1 generation has the heaviest ear weight and good adaptability; using DH351 as the male parent for hybridization with the female parent yields a hybrid with a large ear weight. Figure 3 From the perspective of heterosis groups, hybrids produced from materials of the X-population heterosis group as the male parent have the heaviest ear weight, earlier flowering, and taller plant height; while hybrids produced from materials of the Reid heterosis group as the male parent have the characteristics of high ear weight, short plant height, and early flowering (e.g., ...). Figure 5 ).

[0235] Based on the above research methods, hybrids were classified from the perspectives of paternal parent combinations and heterotic groups, and adaptation analysis diagrams were drawn to systematically evaluate the adaptability of different populations under various environments. By comparing the effects of different paternal parent combinations on key traits of hybrids (such as flowering time, plant height, and ear weight) and analyzing the overall adaptability characteristics of different heterotic groups, a scientific basis is provided for maize breeding under different ecological environments, and breeding strategies are optimized.

[0236] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0237] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A highly efficient crop adaptability assessment method based on genotype-environment interaction effects, characterized in that, Includes the following steps: a. Data acquisition and preprocessing: Collect various phenotypic data of the target crop under multiple environments, remove outliers from the phenotypic data, and perform numerical transformation on the original phenotypic data to make the distribution closer to a normal distribution; b. Environmental mean calculation and centering: Calculate the phenotypic mean of all materials under each environment as a quantitative indicator of the main environmental effect; center the environmental mean to eliminate systematic bias between environments; c. Linear Regression Model Construction: A linear regression model is constructed for each material separately, with the centered environmental mean as the independent variable and the phenotypic value of the material in each environment as the dependent variable. An algorithm is constructed to fit the model. The specific algorithm is as follows: the phenotypic value of a material in a certain environment is equal to the phenotypic value of the material when the environmental mean is zero, i.e., the intercept, plus the product of the material's response strength to environmental changes (i.e., the slope) and the environmental mean, and finally, a residual is added. d. Obtain parameters characterizing adaptability: quantify the average performance of the material by the intercept; the larger the intercept value, the higher the phenotypic level of the material under overall environmental conditions; quantify the plasticity of the material by the slope; the larger the absolute value of the slope, the more sensitive the material is to environmental changes, i.e., the higher the plasticity; plot a two-dimensional scatter plot with the intercept and slope as coordinates, and divide the material into four adaptive regions based on the intercept and slope; e. Model Validation and Output: The goodness of fit of the model is evaluated through residual analysis and the coefficient of determination R² to ensure the accuracy and stability of the evaluation results; Output key parameters, including intercept, slope, and residual information for each material, and generate corresponding visualization charts to intuitively show the performance trends of different materials in multiple environments; f. Adaptive grading: A two-dimensional coordinate system is constructed based on two parameters, intercept and slope, and materials are divided into four categories: general adaptability, stable adaptability, special adaptability and low adaptability, generating a visual chart.

2. The efficient crop adaptability assessment method based on genotype-environment interaction effects according to claim 1, characterized in that: The numerical transformation described in step a specifically involves: Take the absolute value of all phenotypic values, then add 1 to the absolute value, and then perform a logarithmic transformation to base 10 to linearize the data distribution.

3. The efficient crop adaptability assessment method based on genotype-environment interaction effects according to claim 1, characterized in that: The method for dividing the region of the two-dimensional scatter plot in step d includes: The intercept and slope parameters of the material were obtained based on a linear regression model and then standardized. The intercept was standardized using Z-score to ensure its mean was 0; the slope was standardized using a 1-centered method to measure its stability against deviations from 1, thus unifying the reference standard for different properties. Within an environmental range, the overall mean of the intercept was used as the vertical axis boundary, and the mean of the slope as the horizontal axis boundary. Preliminary classification was performed based on the intersection of the fitted line and the mean fitted line (i.e., the slope of 1). For materials that do not intersect, the region is divided into region ① and region ② with the mean intercept value of 0 as the dividing line; 2) For materials that intersect, divide them into region ③ and region ④ with the mean slope of 0 as the dividing line; The materials in each region exhibit specific patterns. Among them, the materials in region ② show broad adaptability and possess the target breeding traits. Therefore, when drawing a two-dimensional scatter plot based on yield traits, for traits that are negatively correlated with yield, the intercept is reversed to ensure that the target traits are classified into region ②.

4. The efficient crop adaptability assessment method based on genotype-environment interaction effects according to claim 1, characterized in that: The four types of adaptive regions in step d are specifically as follows: Region ①: Intercept below 0, that is, when the intercept is below the average performance, it indicates low fitness and poor adaptability to all environments; Region ②: Intercept above 0, i.e., when the intercept is higher than the average performance, it represents broad adaptability and is suitable for all environments; Region ③: The slope is significantly greater than 0, that is, when the slope is significantly greater than the average plasticity, it represents environmental sensitivity and adaptation to favorable conditions; Region ④: The slope is significantly less than 0, that is, when the slope is less than the average plasticity, it represents the stability of adversity, and the potential cannot be realized under favorable conditions; The average performance is the mean of the intercepts of all materials, and the average plasticity is the mean of the slopes of all materials.

5. A quantitative analysis method for crop environmental adaptability based on genotype-environment interaction effects, characterized in that, Quantitative analysis of crop environmental adaptability is performed based on the assessment method described in any one of claims 1-4.

6. The method for quantitative analysis of crop environmental adaptability based on genotype-environment interaction effects according to claim 5, characterized in that, Includes the following steps: Genotype-environment interaction effect decomposition: The phenotypic value is decomposed into the genotype main effect, the environmental main effect, and the genotype-environment interaction effect. The specific process is as follows: The phenotypic value of a crop variety under a specific environment is equal to a pre-set value representing the overall average level, plus the value of the genotype main effect of the crop variety, plus the value of the environmental main effect, plus the value of the genotype-environment interaction effect, and finally plus a residual value to reflect the deviation between the actual measured value and the predicted value obtained through analysis and calculation. The main environmental effects were captured by environmental mean, and the genotype-environment interaction effects were fitted linearly by slope. Calculation of environmental adaptability parameters: The intercept and slope are extracted using a linear regression model to characterize the basic phenotypic value and environmental response intensity of the material, respectively; Phenotypic plasticity is defined as follows: Plasticity is equal to the absolute value of the slope, and the higher the index, the stronger the plasticity of the material. Adaptive dynamic analysis: Constructing a material adaptability matrix by combining intercept and slope parameters; Materials are classified into four categories based on adaptability: broadly adaptable, particularly adaptable, poorly adaptable, and stable adaptable.

7. The method for quantitative analysis of crop environmental adaptability based on genotype-environment interaction effects according to claim 6, characterized in that: The linear fitting method for the genotype-environment interaction effect includes: constructing an algorithm to optimize model parameters. The specific algorithm is: the basic phenotypic value of a material under standard environment is equal to the phenotypic value of the material when the environmental mean is zero, i.e., the intercept, plus the product of the material's response intensity to the environment, i.e., the slope, and the environmental mean, and finally, a residual is added. The significance of the slope was verified by hypothesis testing, and statistically significant environmental adaptation parameters were screened.

8. A method for breeding high-yield and high-quality maize varieties based on environmental adaptability analysis, characterized in that, Includes the following steps: Material screening strategy: From the efficient crop adaptability assessment method based on genotype-environment interaction effect described in claim 4, select the widely adaptable material in region ② as the parent; for the particularly adaptable material in region ③, carry out targeted improvement on its high plasticity traits; Matching materials - environmental analysis: Materials from region ③ are selected for planting in a superior environment to fully realize their phenotypic potential; materials from region ④ are suitable for adverse environments, and their high stability ensures basic yield. Hybrid combination optimization: Under favorable conditions, materials from region ③ are selected as the male parent to maximize hybrid vigor; under adverse conditions, materials from region ④ are selected as the male parent to ensure hybrid stability. Utilization of heterotic groups: Selecting specific paternal parents to optimize target traits based on the adaptive characteristics of heterotic groups; Phenotypic validation and iterative breeding: Verify the adaptability of hybrids through multi-environment phenotypic testing; The parent selection and pairing strategies are iteratively optimized based on the test results.

9. The method for breeding high-yield and high-quality maize varieties based on environmental adaptability analysis according to claim 8, characterized in that: The targeted improvement includes: introducing highly plastic traits into low-fit materials in region ① through gene editing or hybridization; and fixing the stress-stability genes of stable-fit materials in region ④ through phenotypic selection.

Images