Genome prediction method and model for corn genotype and environment cross-modal feature fusion

Abstract

The invention provides a corn genotype and environment cross-modal feature fusion genome prediction method and model, a deep learning method is adopted to model a complex nonlinear transformation relationship between different omics data including a genome and an environment group, and a GECFGP model is constructed to fuse corn cross-genotype and environment interaction features. Time sequence characteristics and important growth periods related to phenotypes are well mined, and the function of predicting the phenotypes of the complex agronomic traits of the corn is achieved. The method shows high robustness in different experimental settings, has superiority in genome prediction of G * E interaction, explores phenotype-related important growth periods and performance of predicting phenotypes across different years, provides a new thought for a genome prediction analysis method, and has wide application prospects. Therefore, theoretical and application value reference is provided for adaptability analysis of corn in different environments and exploration of key growth periods, and development of the field of intelligent breeding is accelerated.

Description

Technical Field

[0001] The present invention belongs to the technical field of genome prediction, and in particular relates to a genome prediction method and model for fusing corn genotype with environmental cross-modal features. Background Art

[0002] As one of the most important food crops in the world, corn plays an irreplaceable role in solving food security problems and meeting the needs of the world's population. However, with the intensification of global climate change, food production is facing unprecedented challenges. Extreme weather and climate instability have increased the uncertainty of food production, posing a major threat to global food security. Therefore, it is of great significance to study the genomic prediction method of genotype-environment interaction (G×E) in corn breeding and production.

[0003] At present, the genomic prediction methods for G×E interactions mainly use classical statistical methods, which can better solve the linear interaction mechanism in G×E interactions, but cannot deeply explore the complex nonlinear interaction mechanism behind G×E interactions. The environmental data of the existing G×E genomic prediction methods mainly use full-growing period EC covariates or sub-growing period EC covariates for feature extraction. This method cannot capture the complex temporal transformation characteristics of the full-growing period or sub-growing period in environmental data. Therefore, the existing methods have limited exploration of the optimal growth period of phenotypes and the prediction accuracy needs to be further improved. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a genome prediction method and model that fuses corn genotype with environmental cross-modal characteristics, so as to predict the phenotype of complex agronomic traits of corn.

[0005] The technical solution adopted by the present invention to solve the above technical problems is: a genome prediction method integrating corn genotype and environmental cross-modal features, comprising the following steps:

[0006] S1: Sequencing maize to obtain whole genome SNP data, and collecting environmental data and field agronomic traits of maize planting sites throughout the growth period;

[0007] S2: Extract and fuse the multidimensional features between sites of the whole genome SNP data from different dimensions including height, width and channel to obtain genotype features;

[0008] S3: Capture and integrate the temporal characteristics of multiple environmental factors of the environmental data of the entire growth period to obtain the temporal environmental characteristics. The specific steps are as follows:

[0009] S31: K-means clustering method was used to group similar environmental factors into the same category;

[0010] S32: Capturing the temporal characteristics of the entire growth period of each environmental factor within the cluster through the autocorrelation mechanism;

[0011] S33: Full growth period temporal characteristics of multiple environmental factors within a class are fused through a fully connected neural network;

[0012] S34: Fusion of inter-class environmental factor features through the cross-attention mechanism;

[0013] S4: The genotype characteristics obtained in step S2 and the temporal environmental characteristics obtained in step S3 are integrated through the Concat mechanism;

[0014] S5: The complex relationships among different omics features are integrated through multi-layer fully connected neural networks; the phenotypic values ​​of specific traits of specific maize materials are predicted and output through the last layer of fully connected neural networks.

[0015] According to the above scheme, in step S1, the specific steps are:

[0016] S11: Whole genome SNP data obtained by sequencing the maize population; random screening of the whole genome SNP data to obtain genotype data G of L SNP sites in the whole genome L ={g1,g2…g L}, g i ∈{0,1,2}, i∈{1,2…L}, the density of the SNP site data after screening is consistent with the density distribution of the data before screening; the SNP site data after screening is converted into a matrix X∈R with a length of L and a number of channels C L×C , R is the field of real numbers;

[0017] S12: Collect environmental factors related to light, temperature, water, air and soil during the whole growth period of the corn planting site; let the i-th environmental factor on the Mth day be x M , the i-th environmental factor within M days is represented as an array f i ={x1,x2…x M}, then the data of f environmental factors in M ​​days is a two-dimensional matrix E∈R f×M ;

[0018] S13: Collect phenotypic values ​​of field agronomic traits including plant height and yield at maize planting sites.

[0019] Furthermore, in step S2, the specific steps are:

[0020] S21: Matrix X∈R constructed for whole genome SNP data L×C Perform the Patch operation to obtain small blocks X (1) ∈R H×C×P, L = H × P, each small block contains P SNP data, and the height of each small block is H;

[0021] Let W embedding is the weight of the Embedding operation, b embedding is the bias of the Embedding operation, D is the dimension after Embedding; for X (1) Through the Embedding operation, the high-dimensional features are mapped to obtain the encoded block embedding feature Y embedding ∈R H×C×D :

[0022] Y embedding =W embedding X (1) +b embedding (1);

[0023] S22: extracting multi-dimensional intra-block features including height, width and channel from the encoded block embedding feature Y, capturing the position information of the block and the feature relationship between the blocks;

[0024] S23: Perform a global average pooling operation on the feature vector Y extracted in step S22. Let Y c,h,d is the element of the cth channel, hth row, and dth column of the input feature vector, then the genotype feature G output by global feature fusion is:

[0025]

[0026] Furthermore, in step S22, the specific steps are:

[0027] S221: Let x i represents the input of the i-th layer neuron, σ i 2 Represents the mean and variance of the input, and ε takes the default value of 1e-5; LayerNorm is performed on the encoded block embedding features to speed up model training and keep the translation and scaling of features unchanged:

[0028]

[0029] S222: Perform channel dimension transformation on the features after layer normalization through Froward Transpose operation, transpose the corresponding dimension to the end, and output it as X (2) ∈R H×D×C ;

[0030] S223: Let W1 and W2 represent the weights of the two-layer linear mapping, b1 and b2 represent the bias of the two-layer linear mapping, σ represents the activation function Gelu, and perform two-layer channel dimension MLP operation on the features after channel dimension transformation to extract the important features X of the channel dimension (3) for:

[0031] X (3) =σ(W2σ(W1X (2) +b1)+b2) (3),

[0032] Among them, the activation function Gelu is used to nonlinearize the captured linear features:

[0033] Gelu(x)=xP(X≤x)X~N(0,1) (4);

[0034] S224: Restore the channel dimension through the Back transpose operation and convert the X output of the MLP operation (3) ∈R H×D×C Adjust to Y (1) ∈R H×C×D As input to the next layer;

[0035] S225: Repeat steps S221 to S224 to extract features of different dimensions including height, width and channel in sequence.

[0036] According to the above scheme, in step S31, the specific steps are:

[0037] S311: Let f j is the jth environmental factor, μ j is the center point of the jth class, and the index c of the cluster to which the jth data point belongs j for:

[0038]

[0039] C i is the set of environmental factors of the i-th category, and the randomized category center μ i for:

[0040]

[0041] S312: Let k be the number of categories, assign the environmental factor to the nearest category, and find the objective function J:

[0042]

[0043] S313: define a new class center as the center of the environmental factors belonging to the class;

[0044] S314: Repeat steps S312 and S313 to minimize the objective function J.

[0045] Furthermore, in step S32, the specific steps are:

[0046] The environmental factor sequence f is transformed into i The linear mapping is query vector Q, key vector K, value vector V, let W q is the weight matrix of the query vector Q, W k is the weight matrix of the key vector K, W v is the weight matrix of the value vector V:

[0047] Q=W q f i (9)

[0048] K=W k f i (10)

[0049] V=W v f i (11);

[0050] Based on the Wiener-Khinchi theory, the time domain is converted to the frequency domain through the fast Fourier transform FFT, and the autocorrelation of the time series is calculated in the frequency domain. * is the conjugate transform of FFT, j is the imaginary unit, n is the index of the median value in the sequence, m is the index of the discrete frequency, and S is the intensity of the different frequency components in the sequence. Q,K for:

[0051]

[0052] The frequency domain is transformed into the time domain through the inverse fast Fourier transform IFFT, and the autocorrelation function R of the time series is obtained. Q,K , the autocorrelation function describes the similarity of the sequence at different time delays, that is, the confidence of the time delay in the sequence:

[0053]

[0054] Select the l time delays with the highest confidence level and take

[0055] τ1,τ2,..,τ l =argTopk(R Q,K (τ)) (14),

[0056] The similarity value of the delay is normalized by the softmax function to obtain the weight of the delay:

[0057]

[0058] V(t-τ i ) is V at τ i After the delay, the rolling sequence is weighted and summed up, and the features of the subsequences with different delays are aggregated to obtain the final result of feature extraction, Auto-Correlation:

[0059]

[0060] Furthermore, in step S34, the specific steps are:

[0061] Learn the correlation between environmental factors and perform weighted fusion of features of different categories according to the learned weights to extract inter-class feature information;

[0062] For k categories, a cross-attention mechanism is used to extract features between each two to obtain k*(k-1) feature vectors e:

[0063] CrossAttention(Q,K,V)=SoftMax(Q·K T )·V (17),

[0064] e i =CrossAttention(Q i ,K i+1 ,V i+1 ),i∈{1,2,...,n} (18), These feature vectors are fused through the Concat layer to obtain the temporal environment feature E:

[0065] E=Concat(e1,e2,…,e k*(k-1) ) (19).

[0066] Furthermore, in step S4, the specific steps are:

[0067] The Concat mechanism is used to fuse the genotype feature G extracted in step S2 and the temporal environmental feature E extracted in step S3:

[0068] F = Concat(G, E) (20).

[0069] Furthermore, in step S5, the specific steps are:

[0070] Let W MLP1 , W MLP2 and W MLP3 are the weights of the three-layer fully connected neural network, b MLP1 、b MLP2 and b MLP3are the biases of the three-layer fully connected neural network respectively. The fusion feature F obtained in step S4 is fused with the complex relationship between different omics features through a multi-layer fully connected neural network. The phenotypic value Phe of the specific trait of the specific material of corn is predicted and output through the last layer of the fully connected neural network:

[0071] Phe=W MLP3 (W MLP2 BatchNorm(W MLP1 F+b MLP1 )+b MLP2 )+b MLP3 (twenty one);

[0072] Let mean[x], var[x] be the global mean and variance obtained during network training. The default value of ε is 1e-05. γ is a scaling factor that allows the model to recover some information lost during normalization. β is a translation factor that allows the model to adjust the distribution of inputs. Dropout and batch normalization BatchNorm operations are used to prevent gradient vanishing and overfitting while capturing nonlinear features:

[0073]

[0074] Let N represent the number of materials involved in the MSELoss calculation, and use the mean of the sum of squares of the difference between the predicted phenotypic value Phe and the true phenotypic value y to measure the network loss, and the mean square error loss function MSELoss is:

[0075]

[0076] A genomic prediction model for the fusion of maize genotype and environmental cross-modal features is constructed based on a genomic prediction method for the fusion of maize genotype and environmental cross-modal features;

[0077] It includes genotype feature extraction module, temporal environmental feature extraction module and cross-omics feature fusion module;

[0078] The genotype feature extraction module is a fully connected neural network module MDMLP for multi-dimensional spatial feature fusion, including Patch Embedding layer, N MDblock layers, global average pooling layer and fully connected layer, which is used to extract and fuse the multi-dimensional features of SNP site data; Patch Embedding layer includes Patch layer and Embedding layer; Patch layer is used for block division; Embedding layer is used for encoding; each MDblock layer includes token mixing layer Base-dim MDLayer, height mixing layer Height MDlayer and channel mixing layer Channel MDlayer, which are used to extract input features in different dimensions respectively; global average pooling layer is used for global feature fusion, which improves the model's ability to represent global SNP features and reduces the impact of position bias on feature extraction;

[0079] The temporal environmental feature extraction module first uses the K-means clustering method to cluster environmental factors into k categories, then captures the temporal features of the entire growth period of each environmental factor within the category through the autocorrelation mechanism, then fuses the full growth period features of multiple environmental factors within the category through a fully connected neural network, and finally realizes the fusion of environmental factor features between categories through the cross-attention mechanism, thereby capturing the temporal features of the entire growth period of multiple environmental factors;

[0080] The cross-omics feature fusion module regards each omics as a modality and uses the Concat mechanism to construct a cross-omics feature fusion layer to effectively fuse the captured genotypes with environmental temporal characteristics and output phenotypic values ​​for predicting specific traits of corn.

[0081] The beneficial effects of the present invention are:

[0082] 1. The present invention provides a genomic prediction method and model for fusing cross-modal features of corn genotype and environment, which adopts a deep learning method to model the complex nonlinear transformation relationship between different omics data including genome and environment groups, and integrates the characteristics of corn cross-genotype and environment interaction by constructing a GECF_GP model, thereby better mining the temporal characteristics and important growth periods related to the phenotype, and realizing the function of predicting the phenotype of complex agronomic traits of corn.

[0083] 2. The present invention demonstrates strong robustness in different experimental settings. Based on the maize976 corn population dataset, the patterns of untested genotypes in observed environments and tested genotypes in unobserved environments verify that the accuracy of the present invention in predicting field agronomic trait phenotypes is better than the existing six cutting-edge genomic prediction methods, and is superior in genomic prediction of G×E interactions, providing new ideas for genomic prediction analysis methods, and thus providing a theoretical reference for the adaptability analysis of corn in different environments.

[0084] 3. The present invention explores the important growth periods related to phenotypes and the performance of predicting phenotypes across different years. In terms of important growth periods, the present invention has significant differences in the predicted phenotypic distributions for the entire growth period and the six sub-growth periods, and the prediction accuracy for the sub-growth period is slightly higher than that for the entire growth period; the model trained on a specific genotype in an unobserved environment is more suitable for realizing phenotypic prediction across years than the model trained on an untested genotype in an observed environment. It provides a reference for application value in the exploration of the key growth period of corn and accelerates the development of the field of smart breeding. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 is a flow chart of an embodiment of the present invention.

[0086] Figure 2 It is a hierarchical framework diagram of the GECF_GP model of an embodiment of the present invention.

[0087] Figure 3 It is a comparison chart of the performance of the embodiment of the present invention and different existing methods for predicting the Maize976 population in the observed environment for untested genotypes.

[0088] Figure 4 It is a comparison chart of the performance of the embodiment of the present invention and different existing methods for predicting the Maize976 population in an unobserved environment using the test genotype.

[0089] Figure 5 It is a performance evaluation diagram of the embodiment of the present invention for predicting the phenotype of the Maize976 population during the whole growth period and six sub-growth periods.

[0090] Figure 6 It is a performance evaluation diagram of a model trained with an untested genotype in an observed environment and a model trained with a known genotype in an unobserved environment in predicting phenotypes across years according to an embodiment of the present invention. DETAILED DESCRIPTION

[0091] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0092] Example 1

[0093] See also Figure 1 , this embodiment includes the following steps:

[0094] S1: Sequencing to obtain single nucleotide genetic variation site (SNP) data of the whole corn genome, collecting environmental data of each planting site throughout the growth period and field agronomic traits of each planting site;

[0095] The single nucleotide variation data (SNP) of the whole genome of the maize population obtained by sequencing was randomly screened to obtain the genotype data G of L SNP sites in the whole genome. L ={g1,g2…g L}, g i ∈{0,1,2}, i∈{1,2…L}, the density of SNP sites after screening is consistent with the density distribution before screening; convert it into a matrix X∈R with a length of L and a number of channels C L×C , R is the field of real numbers;

[0096] For environmental data, collect the environmental factors related to light, temperature, water, air and soil every day during the corn planting period; let the i-th environmental factor on the Mth day be x M , the i-th environmental factor within M days is represented as an array f i ={x1,x2…x M}, then the data of f environmental factors in M ​​days is a two-dimensional matrix E∈R f×M .

[0097] For the field agronomic trait data of each breeding material, the phenotypic values ​​of agronomic traits such as plant height and yield were collected at each planting location.

[0098] S2: In the GECF_GP model, for the input whole genome SNP data, the fully connected neural network (MDMLP) module with multidimensional spatial feature fusion is used to effectively extract the complex relationship features between SNP sites from different dimensions (height, width and channel);

[0099] MDMLP includes Patch Embedding, MDblock, global average pooling and fully connected layers, which are used to extract and fuse the multi-dimensional features of SNP locus data. The specific implementation steps are as follows:

[0100] S201: Matrix X∈R constructed from SNP data L×C Input to the Patch Embedding layer, use Patch to divide X into blocks to get X (1) ∈R H×C×P , where L = H × P; each small block contains P SNP data, and the height of each small block is H. (1) The high-dimensional features are mapped through the Embedding layer to obtain the encoded block embedding feature Y embedding , the specific implementation process is shown in formula (1).

[0101] Y embedding =W embedding X (1) +b embedding (1)

[0102] In formula (1), W embedding is the weight of the embedding layer, b embedding is the bias of the embedding layer, D is the dimension after embedding, and Y embedding ∈R H×C×D .

[0103] S202: The encoded block embedding features are input into the MDblock layer for feature extraction, where each MDblock consists of three MDLayers. The three MDLayer layers represent the input features for feature extraction in different dimensions, including the token mixing layer (Base-dim MDLayer), the height mixing layer (Height MDlayer) and the channel mixing layer (Channel MDlayer). By applying MLP in multiple dimensions, not only the intra-block features can be captured, but also the location information of the blocks and the feature relationship between the blocks can be captured.

[0104] First, the feature vector of the input MDLayer is subjected to layer normalization (LayerNorm) to speed up the model training and keep the translation and scaling of the features unchanged. The implementation process is shown in formula (2).

[0105]

[0106] In formula (2), x i represents the input of the i-th layer neuron, σ i 2 Represents the mean and variance of the input, and ε takes the default value of 1e-5.

[0107] Secondly, for the features after layer normalization, the input is dimensionally changed to facilitate operations on the corresponding dimensions. Specifically, the input data is represented by Y∈R H×C×D After the output of the Embedding layer and layer normalization, the data will be re-arranged by the Froward Transpose layer, and the corresponding dimension will be transposed to the end. H×C×D , the output is X (2) ∈R H×D×C .

[0108] Furthermore, the features after channel dimension transformation are subjected to two layers of channel dimension MLP operation to extract important features of the channel dimension. The specific implementation process is shown in formula (3).

[0109] X (3) =σ(W2σ(W1X (2) +b1)+b2) (3)

[0110] In formula (3), W1 and W2 represent the weights of the two-layer linear mapping, b1 and b2 represent the biases of the two-layer linear mapping, and σ represents the activation function Gelu. This operation can nonlinearize the captured linear features. The specific implementation process is shown in formula (4).

[0111] Gelu(x)=xP(X≤x) X~N(0,1) (4)

[0112] Finally, the channel dimension is restored by back transpose, and the output X of MLP is (3) ∈R H×D×C Adjust to Y (1) ∈R H×C×D as input to the next layer.

[0113] S203: Repeat the process of S202 above to extract multi-dimensional features of the token mixing layer, the height mixing layer and the channel mixing layer in turn.

[0114] S204: Input the feature vector Y extracted by the MDBlock layer into the global average pooling layer to achieve global feature fusion, improve the model's ability to represent global SNP features, and reduce the impact of position bias on feature extraction. The specific implementation process is shown in formula (5).

[0115]

[0116] In formula (5), Y c,h,d represents the element of the cth channel, hth row, and dth column of the input vector, and G represents the output feature after global average pooling, that is, the extracted SNP feature.

[0117] S3: For the input full-growing period environmental data, the K-means clustering method is first used to cluster the environmental factors into k categories. Secondly, the full-growing period temporal characteristics of each environmental factor in the class are captured through the autocorrelation mechanism. The full-growing period characteristics of multiple environmental factors in the class are then fused through a fully connected neural network. Finally, the cross-attention mechanism is used to achieve the fusion of environmental factor characteristics between classes, thereby capturing the temporal characteristics of multiple environmental factors in the full growing period.

[0118] S301: Cluster the environmental factors into k categories using the K-means clustering method. Similar environmental factors are classified into the same category to facilitate further analysis and processing.

[0119] The K-means method consists of the following steps:

[0120] First, randomize the cluster center μ i, and then assign the environmental factor to the nearest class, and finally define the new class center as the center of the environmental factor belonging to this class, and repeat steps 2 and 3 to minimize the objective function J. The formal definitions of the above steps are shown in formulas (6), (7) and (8).

[0121]

[0122] In formula (6), f j refers to the jth environmental factor, μ j represents the center point of the jth class, where c j represents the index of the cluster to which the jth data point belongs. In formula (7), C i represents the set of the i-th type of environmental factors, and k is the number of categories in formula (8).

[0123] S302: Capture the full-growing period time series characteristics of each environmental factor in the class through the autocorrelation mechanism. The autocorrelation mechanism is a further optimization of the shortcomings of the self-attention mechanism in time series data processing. The mechanism calculates the similarity between sequences based on the relationship of periodic dependency and aggregates similar subsequences through time delay information. The autocorrelation mechanism uses its inherent sparsity and subsequence-level similarity for characterization, thereby improving computational efficiency and information utilization. The autocorrelation mechanism converts the environmental factor sequence f i Perform linear mapping to query vector Q, key vector K, value vector V, let W q is the weight matrix of the query vector Q, W k is the weight matrix of the key vector K, W v is the weight matrix of the value vector V:

[0124] Q=W q f i (9)

[0125] K=W k f i (10)

[0126] V=W v f i (11)

[0127] Based on the Wiener-Khinchi theory, the time domain is converted to the frequency domain through the fast Fourier transform, and the autocorrelation of a time series can be efficiently calculated in the frequency domain. Then, the frequency domain is transformed to the time domain through the inverse Fourier transform (IFFT) to obtain the autocorrelation function of the time series. The autocorrelation function describes the similarity of the sequence at different time delays, that is, the confidence of the time delay in the sequence. The specific implementation process is shown in formulas (12) and (13).

[0128]

[0129] In formula (12), FFT refers to fast Fourier transform. * represents its conjugate transformation, S Q,K represents the intensity of different frequency components in the sequence, j is an imaginary unit, n is the index of the median in the sequence, m is the index of the discrete frequency, and in formula (13) IFFT refers to the inverse fast Fourier transform, R Q,K is the similarity of the sequence to itself at different time delays.

[0130] Select the l time delays with the highest confidence, normalize their similarity values ​​through the softmax function to obtain the weight of the time delay, further weight the sum of similar subsequences corresponding to the time delay, and then aggregate the features of subsequences with different time delays to obtain the final result Auto-Correlation after feature extraction. Therefore, the time complexity of the autocorrelation mechanism is O(DlogD), which is different from the self-attention mechanism O(D 2 ) The time complexity has been greatly improved. The specific process is shown in formula (14), formula (15) and formula (16).

[0131] τ1,τ2,..,τ l =argTopk(R Q,K (τ)) (14)

[0132]

[0133] In formula (16), V(t-τ i ) indicates that V is in τ i Delayed scrolling sequence.

[0134] S303: using a fully connected neural network to fuse the temporal characteristics of the entire growth period of multiple environmental factors extracted within the class;

[0135] S304: The cross-attention mechanism is used to achieve the fusion of inter-class environmental factor features, thereby capturing the temporal characteristics of multiple environmental factors throughout the entire growth period. This mechanism aims to learn the correlation between environmental factor classes and perform weighted fusion of features of different categories based on the learned weights, thereby extracting inter-class feature information to enhance clustering effects and feature expression capabilities.

[0136] Specifically, this method maps the feature vector of one category to vector Q through a linear layer, and maps the feature vector of the other category to K and V. The specific implementation process is shown in formulas (9), (10) and (11). The specific process of cross-attention is shown in formula (17).

[0137] CrossAttention(Q,K,V)=SoftMax(Q·K T )·V (17)

[0138] For k categories, after extracting features through the cross-attention mechanism between each two, k*(k-1) feature vectors e can be obtained. Subsequently, these feature vectors are fused through the Concat layer to obtain the final environment feature vector E. The specific implementation process is shown in formulas (18) and (19).

[0139] e i =CrossAttention(Q i ,K i+1 ,V i+1 ),i∈{1,2,...,n} (18)

[0140] E=Concat(e1,e2,…,e k*(k-1) ) (19)

[0141] S4: The Concat mechanism is used to construct a cross-omics feature fusion layer to effectively fuse the captured genotype features with temporal environmental features;

[0142] Based on the genotype feature G extracted in step S2 and the temporal environmental feature E fused with multiple environmental factors extracted in step S3, the Concat mechanism is used to achieve the fusion of cross-genomic and environmental omics features. The specific implementation is shown in formula (20).

[0143] F=Concat(G,E) (20)

[0144] S5: The output layer uses a multi-layer fully connected neural network to further integrate the complex relationships between different omics features. Among them, the last layer of fully connected neural network is used to predict the phenotypic value of a specific corn material.

[0145] Let W MLP1 , W MLP2 and W MLP3 are the weights of the three-layer fully connected neural network, b MLP1 、b MLP2 and b MLP3 are the biases of the three-layer fully connected neural network respectively. The fusion feature F obtained in step S4 is further integrated through the multi-layer fully connected neural network to strengthen the complex relationship between different omics features. In addition, the last layer of fully connected neural network is used to predict the phenotypic value Phe of the specific trait of the specific maize material. The specific implementation is shown in formula (21).

[0146] Phe=W MLP3 (W MLP2 BatchNorm(WMLP1 F+b MLP1 )+b MLP2 )+b MLP3 (twenty one)

[0147] In the process of network training, in order to prevent gradient disappearance and overfitting and capture nonlinear features at the same time, dropout and batch normalization BatchNorm operations are used in the network. The batch normalization implementation process is shown in formula (22). Dropout is another commonly used regularization technique that can randomly set the output of some neurons in the network to zero to reduce the dependency between neurons and prevent overfitting.

[0148]

[0149] In formula (22), mean[x], var[x] are the global mean and variance obtained during network training; the default value of ε is 1e-05; γ is a scaling factor that allows the model to recover some information lost during normalization; β is a translation factor that allows the model to adjust the distribution of the input.

[0150] When calculating the network loss, the mean of the sum of squares of the differences between the predicted phenotypic value Phe and the true phenotypic value y of the material is used for measurement, that is, the mean square error loss function MSELoss, and the specific implementation is shown in formula (23).

[0151]

[0152] In formula (23), N represents the number of materials involved in the MSELoss calculation.

[0153] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0154] Example 2

[0155] See also Figure 2 In this embodiment, a genomic prediction model GECF_GP (genotype environment cross-modal feature fusion genetic prediction) for fusion of maize genotype and environmental cross-modal features is constructed through the steps of Example 1, including a genotype feature extraction module, a temporal environmental feature extraction module, and a cross-omics (each omics is regarded as a modality) feature fusion module.

[0156] The genotype feature extraction module effectively extracts the complex relationship features between SNP sites from different dimensions (height, width and channel) through the fully connected neural network (MDMLP) module with multidimensional spatial feature fusion.

[0157] The temporal environmental feature extraction module first uses the K-means clustering method to cluster environmental factors into k categories, then captures the temporal features of the entire growth period of each environmental factor within the class through the autocorrelation mechanism, and then fuses the full growth period features of multiple environmental factors within the class through a fully connected neural network. Finally, the cross-attention mechanism is used to achieve the fusion of environmental factor features between classes, thereby capturing the temporal features of the entire growth period of multiple environmental factors.

[0158] The cross-omics feature fusion module uses the Concat mechanism to construct a cross-omics feature fusion layer to effectively fuse the captured genotypes with environmental temporal features.

[0159] Example 3

[0160] Based on the steps of Example 1, this example performs genome prediction on the genotype data of the Maize976 corn population, the environmental data of 11 environments planted in 2011 and 2012, and the field agronomic trait data, and evaluates the prediction results as follows:

[0161] The first study evaluated the phenotypic prediction performance of untested genotypes in an observed environment. The phenotypic prediction accuracy of GECF_GP was compared with that of six cutting-edge G×E genomic prediction methods in the Maize976 population. Figure 3 The six G×E genomic prediction methods compared included the main effect EAD plus GE bias model based on the best linear unbiased prediction of the genome (EAD+GE(GB)), the main effect EAW plus GE reaction norm model based on the best linear unbiased prediction of the genome (EADW+GW(GB)), the main effect EAD plus GE bias model based on Gaussian kernel (EAD+GE(GK)), the main effect EAW plus GE reaction norm model based on Gaussian kernel EADW+GW(GK), the main effect EAD plus GE bias model based on deep kernel EAD+GE(DK), and the main effect EAW plus GE reaction norm model based on deep kernel EADW+GW(DK)).

[0162] For plant height (PH), the prediction accuracy of GECF_GP ranged from 0.64 to 0.87 in 11 environments, while the prediction accuracy of the other six statistical methods ranged from 0.489 to 0.838. The prediction accuracy of all methods was low in GX2011, HB2011, and SC2011, but high in DHN2011, CQ2011, and HN2011. In each environment, the accuracy of GECF_GP was at least 3% higher than that of all the compared methods. Overall, the average accuracy of GECF_GP was 0.7744, which was better than all the compared methods. For yield (EW), the prediction accuracy of GECF_GP ranged from 0.296 to 0.5154 in 7 environments, while the prediction accuracy of the other six statistical methods ranged from 0.113 to 0.481. In each environment, the environments with higher prediction accuracy are DHN2011 and YN2012, while the environments with lower prediction accuracy are CQ2012, HN2012 and YN2011. In the prediction of EW traits, the prediction accuracy of GECF_GP in each environment is also higher than that of all the comparison methods. Among the comparison methods, EAD+GE(DK) still has the highest average prediction accuracy of 0.3915, which is slightly lower than GECF_GP. The worst prediction accuracy is EADW+GW(GB), with an average accuracy of 0.3317, which is significantly lower than GECF_GP.

[0163] The second evaluation tested the phenotype prediction performance of the tested genotypes in unobserved environments. The prediction accuracy of GECF_GP was compared with that of six cutting-edge G×E genomic prediction methods in the Maize976 population (e.g. Figure 4 shown).

[0164] In the prediction of PH traits, GECF_GP has the best prediction performance. In the 2011 and 2012 environments, the average prediction accuracy of GECF_GP was 1% and 0.6% higher than the best comparison method, respectively, which was slightly higher than all the comparison methods. Although the accuracy of GECF_GP was 3% lower than that of EAD+GE(DK) in GX2011 in 2011, and the accuracy of GECF_GP (0.873) was slightly lower than that of EADW+GW(GB) (0.874) in HN2012 in 2012, the overall average accuracy of GECF_GP was still higher than that of all the comparison methods. In the prediction of EW traits, the six cutting-edge G×E genomic prediction methods all showed obvious instability, and each method showed negative correlation in some environmental tests. For example, in the DHN environment in 2011, the prediction accuracy of EADW+GW(GK) was -0.3219. Although only in a few environments (such as HN2011 and HN2012), the prediction accuracy of GECF_GP is higher than that of other comparison methods. However, the average accuracy of GECF_GP in the 2011 and 2012 environments is 0.458 and 0.425, respectively, which is 12.2% and 7.99% higher than the best comparison methods (EAD+GE(GK) and EADW+GW(GK)), respectively, showing a significant advantage. In addition, GECF_GP ranks among the top in all environments, and there is no negative correlation, indicating that GECF_GP is more stable than the comparison methods.

[0165] The third project evaluated the accuracy of GECF_GP in predicting pH under the observed environment for untested genotypes in the full growth period (0-150 days) and six sub-growth periods (0-15 days, 16-35 days, 36-65 days, 66-90 days, 91-120 days, and 121-150 days). The experimental results are shown in Figure 5 shown.

[0166] In the 11 environments, the accuracy of PH prediction in the whole growth period and the six sub-growth periods of each environment was not much different, but the phenotypic distribution predicted by the whole growth period and the sub-growth period was significantly different (p-value<0.05). In most environments, the third growth period (36-65 days) was closest to the prediction results of the whole growth period, followed by the second growth period (16-35 days) and the fourth growth period (66-90 days). The above results have important reference value for practical breeding applications, indicating that using only the environmental data of these growth periods for phenotypic prediction can obtain results similar to those of the whole growth period, thereby saving breeding resources.

[0167] The fourth item evaluates the performance of GECF_GP in predicting PH and EW in different years at the same location.

[0168] In order to evaluate the performance of cross-year prediction, two models were designed for cross-year phenotypic prediction. M1 was based on the model constructed by untested genotypes in the observed environment to predict the phenotypic data of 2012 across years, and M2 was based on the model constructed by tested genotypes in the untested environment to predict the phenotypic data of 2012 across years. The experimental results are shown in Figure 2. Figure 6 As shown in the figure, the prediction accuracy of the M2 evaluation method is significantly higher than that of the M1 evaluation method. In the prediction results of PH traits, the accuracy of CQ, HB, HN and YN environments increased by 4.86%, 14.52%, 4.83% and 6.68%, respectively. In the prediction of EW traits, the accuracy of HN and YN environments increased by 12.6% and 12.1%, respectively. When the environmental data set of the training set includes more environmental data, this method performs better in cross-environment prediction. It can be obtained that the model (M2) trained with known genotypes in unobserved environments is more suitable for cross-year phenotypic prediction than the model (M1) trained with untested genotypes in observed environments.

[0169] The above embodiments are only used to illustrate the design ideas and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The protection scope of the present invention is not limited to the above embodiments. Therefore, any equivalent changes or modifications made based on the principles and design ideas disclosed by the present invention are within the protection scope of the present invention.

Claims

1. A genome prediction method for maize genotype and environmental cross-modal features, characterized by: The following steps are involved: S1: Sequencing maize to obtain whole genome SNP data, and collecting environmental data and field agronomic traits of maize planting sites throughout the growth period; S2: Extract and fuse the multidimensional features between sites of the whole genome SNP data from different dimensions including height, width and channel to obtain genotype features; S3: Capture and integrate the temporal characteristics of multiple environmental factors of the environmental data of the entire growth period to obtain the temporal environmental characteristics. The specific steps are as follows: S31: K-means clustering method was used to group similar environmental factors into the same category; S32: Capturing the temporal characteristics of the entire growth period of each environmental factor within the cluster through the autocorrelation mechanism; S33: Full growth period temporal characteristics of multiple environmental factors within a class are fused through a fully connected neural network; S34: Fusion of inter-class environmental factor features through the cross-attention mechanism; S4: The genotype characteristics obtained in step S2 and the temporal environmental characteristics obtained in step S3 are integrated through the Concat mechanism; S5: The complex relationships among different omics features are integrated through multi-layer fully connected neural networks; the phenotypic values ​​of specific traits of specific maize materials are predicted and output through the last layer of fully connected neural networks.

2. The method for genome prediction by fusing corn genotype with cross-modal environmental features according to claim 1, characterized in that: In the step S1, the specific steps are: S11: Whole genome SNP data obtained by sequencing the maize population; random screening of the whole genome SNP data to obtain genotype data G of L SNP sites in the whole genome L ={g1,g2…g L }, g i ∈{0,1,2}, i∈{1,2…L}, the density of the SNP site data after screening is consistent with the density distribution of the data before screening; the SNP site data after screening is converted into a matrix X∈R with a length of L and a number of channels C L×C , R is the field of real numbers; S12: Collect environmental factors related to light, temperature, water, air and soil during the whole growth period of the corn planting site; let the i-th environmental factor on the Mth day be x M , the i-th environmental factor within M days is represented as an array f i ={x1,x2…x M }, then the data of f environmental factors in M ​​days is a two-dimensional matrix E∈R f×M ; S13: Collect phenotypic values ​​of field agronomic traits including plant height and yield at maize planting sites.

3. The genome prediction method of corn genotype and environmental cross-modal feature fusion according to claim 2, characterized in that: In the step S2, the specific steps are: S21: Matrix X∈R constructed for whole genome SNP data L×C Perform the Patch operation to obtain small blocks X (1) ∈R H ×C×P , L = H × P, each small block contains P SNP data, and the height of each small block is H; Let W embedding is the weight of the Embedding operation, b embedding is the bias of the Embedding operation, D is the dimension after Embedding; for X (1) Through the Embedding operation, the high-dimensional features are mapped to obtain the encoded block embedding feature Y embedding ∈R H×C×D : Y embedding =W embedding X (1) +b embedding (1); S22: extracting multi-dimensional intra-block features including height, width and channel from the encoded block embedding feature Y, capturing the position information of the block and the feature relationship between the blocks; S23: Perform a global average pooling operation on the feature vector Y extracted in step S22. Let Y c,h,d is the element of the cth channel, hth row, and dth column of the input feature vector, then the genotype feature G output by global feature fusion is:

4. The genome prediction method of corn genotype and environmental cross-modal feature fusion according to claim 3, characterized in that: In the step S22, the specific steps are: S221: Let x i represents the input of the i-th layer neuron, σ i 2 Represents the mean and variance of the input, and ε takes the default value of 1e-5; LayerNorm is performed on the encoded block embedding features to speed up model training and keep the translation and scaling of features unchanged: S222: Perform channel dimension transformation on the features after layer normalization through Froward Transpose operation, transpose the corresponding dimension to the end, and output it as X (2) ∈R H×D×C ; S223: Let W1 and W2 represent the weights of the two-layer linear mapping, b1 and b2 represent the bias of the two-layer linear mapping, σ represents the activation function Gelu, and perform two-layer channel dimension MLP operation on the features after channel dimension transformation to extract the important features X of the channel dimension (3) for: X (3) =σ(W2σ(W1X (2) +b1)+b2) (3), Among them, the activation function Gelu is used to nonlinearize the captured linear features: Gelu(x)=xP(X≤x)X~N(0,1) (4); S224: Restore the channel dimension through the Back transpose operation and convert the X output by the MLP operation (3) ∈R H×D×C Adjust to Y (1) ∈R H×C×D As input to the next layer; S225: Repeat steps S221 to S224 to extract features of different dimensions including height, width and channel in sequence.

5. The genome prediction method of corn genotype and environmental cross-modal feature fusion according to claim 1, characterized in that: In the step S31, the specific steps are: S311: Let f j is the jth environmental factor, μ j is the center point of the jth class, and the index c of the cluster to which the jth data point belongs j for: C i is the set of environmental factors of the i-th category, and the randomized category center μ i for: S312: Let k be the number of categories, assign the environmental factor to the nearest category, and find the objective function J: S313: define a new class center as the center of the environmental factors belonging to the class; S314: Repeat steps S312 and S313 to minimize the objective function J.

6. The genome prediction method of corn genotype and environmental cross-modal feature fusion according to claim 5, characterized in that: In the step S32, the specific steps are: The environmental factor sequence f is transformed into i The linear mapping is query vector Q, key vector K, value vector V, let W q is the weight matrix of the query vector Q, W k is the weight matrix of the key vector K, W v is the weight matrix of the value vector V: Q=W q f i (9), K=W k f i (10), V=W v f i (11); Based on the Wiener-Khinchi theory, the time domain is converted to the frequency domain through the fast Fourier transform FFT, and the autocorrelation of the time series is calculated in the frequency domain. * is the conjugate transform of FFT, j is the imaginary unit, n is the index of the median value in the sequence, m is the index of the discrete frequency, and S is the intensity of the different frequency components in the sequence. Q,K for: The frequency domain is transformed into the time domain through the inverse fast Fourier transform IFFT, and the autocorrelation function R of the time series is obtained. Q,K , the autocorrelation function describes the similarity of the sequence at different time delays, that is, the confidence of the time delay in the sequence: Select the l time delays with the highest confidence level and take τ1,τ2,..,τ l =argTopk(R Q,K (t)) (14), The similarity value of the delay is normalized by the softmax function to obtain the weight of the delay: V(t-τ i ) is V at τ i After the delay, the rolling sequence is weighted and summed up, and the features of the subsequences with different delays are aggregated to obtain the final result of feature extraction, Auto-Correlation:

7. The genome prediction method of corn genotype and environmental cross-modal feature fusion according to claim 6, characterized in that: In the step S34, the specific steps are: Learn the correlation between environmental factors and perform weighted fusion of features of different categories according to the learned weights to extract inter-class feature information; For k categories, a cross-attention mechanism is used to extract features between each two to obtain k*(k-1) feature vectors e: CrossAttention(Q,K,V)=SoftMax(Q·K T )·V (17), e i =CrossAttention(Q i ,K i+1 ,V i+1 ),i∈{1,2,...,n} (18), These feature vectors are fused through the Concat layer to obtain the temporal environment feature E: E=Concat(e1,e2,...,e k*(k-1) ) (19)。 8. The genome prediction method of corn genotype and environmental cross-modal feature fusion according to claim 3 or 7, characterized in that: In the step S4, the specific steps are: The Concat mechanism is used to fuse the genotype feature G extracted in step S2 and the temporal environmental feature E extracted in step S3: F = Concat(G, E) (20).

9. The genome prediction method of corn genotype and environmental cross-modal feature fusion according to claim 8, characterized in that: In the step S5, the specific steps are: Let W MLP1 , W MLP2 and W MLP3 are the weights of the three-layer fully connected neural network, b MLP1 、b MLP2 and b MLP3 are the biases of the three-layer fully connected neural network respectively. The fusion feature F obtained in step S4 is fused with the complex relationship between different omics features through a multi-layer fully connected neural network. The phenotypic value Phe of the specific trait of the specific material of corn is predicted and output through the last layer of the fully connected neural network: Phe=W MLP3 (W MLP2 BatchNorm(W MLP1 F+b MLP1 )+b MLP2 )+b MLP3 (21); Let mean[x], var[x] be the global mean and variance obtained during network training, and the default value of ε is 1e-05; γ is a scaling factor that allows the model to recover some information lost during normalization; β is a translation factor that allows the model to adjust the distribution of the input; dropout and batch normalization BatchNorm operations are used to prevent gradient vanishing and overfitting while capturing nonlinear features: Let N represent the number of materials involved in the MSELoss calculation, and use the mean of the sum of squares of the difference between the predicted phenotypic value Phe and the true phenotypic value y to measure the network loss, and the mean square error loss function MSELoss is:

10. A genome prediction model integrating maize genotype and environmental cross-modal features, characterized by: A genome prediction method based on the fusion of corn genotype and environmental cross-modal features according to any one of claims 1 to 9 is constructed; It includes genotype feature extraction module, temporal environmental feature extraction module and cross-omics feature fusion module; The genotype feature extraction module is a fully connected neural network module MDMLP for multi-dimensional spatial feature fusion, including a Patch Embedding layer, N MDblock layers, a global average pooling layer, and a fully connected layer, which is used to extract and fuse the multi-dimensional features of SNP site data; The Patch Embedding layer includes the Patch layer and the Embedding layer; the Patch layer is used for block division; the Embedding layer is used for encoding; each MDblock layer includes the token mixing layer Base-dim MDLayer, the height mixing layer Height MDlayer and the channel mixing layer Channel MDlayer, which are used to extract input features in different dimensions respectively; the global average pooling layer is used for global feature fusion, improving the model's ability to represent global SNP features and reducing the impact of position bias on feature extraction; The temporal environmental feature extraction module first uses the K-means clustering method to cluster environmental factors into k categories, then captures the temporal features of the entire growth period of each environmental factor within the category through the autocorrelation mechanism, then fuses the full growth period features of multiple environmental factors within the category through a fully connected neural network, and finally realizes the fusion of environmental factor features between categories through the cross-attention mechanism, thereby capturing the temporal features of the entire growth period of multiple environmental factors; The cross-omics feature fusion module regards each omics as a modality and uses the Concat mechanism to construct a cross-omics feature fusion layer to effectively fuse the captured genotypes with environmental temporal characteristics and output phenotypic values ​​for predicting specific traits of corn.

Images