Protein interaction prediction method based on hybrid graph attention neural network

Abstract

The invention discloses a protein interaction prediction method based on a mixed graph attention neural network, and belongs to the technical field of protein interaction prediction, and the method comprises the following steps: preprocessing protein data; extracting and fusing the characteristics of the protein; constructing a TCN neural network; constructing a hybrid graph attention neural network; model training; and predicting the interaction between the proteins. According to the method, the two features are fused as protein feature expression, the protein global features are extracted by adopting a TCN network, and the global features do not need to be extracted through a convolutional neural network; the local features of the protein are extracted by sampling GAT and GCN networks, and the local features do not need to be independently extracted through a sliding window technology.

Description

Technical Field

[0001] The present invention relates to the technical field of protein interaction prediction, and in particular to a protein interaction prediction method based on a hybrid graph attention neural network. Background Art

[0002] Protein-Protein Interaction (PPI) is a very important research topic in biology. It refers to the functional connection between two or more protein molecules inside and outside the cell through specific interactions. PPI relationships are crucial for understanding cell functions, signal transduction, metabolic processes, etc. Interactions between proteins are the basis of intracellular biological processes, and they play a key role in a variety of biological activities, including gene expression regulation, cell cycle, immune response, signal transduction, etc. The study of protein interactions is crucial for a deep understanding of the role of proteins in various biological processes in organisms. This will not only help promote disease treatment and drug development, but also promote progress in the medical field and deepen the understanding of molecular mechanisms. However, the current prediction of protein-protein interactions mainly relies on laboratory methods, which are usually time-consuming and labor-intensive, require a lot of resources, and can only be applied to specific proteins, and are difficult to be universally applicable to all proteins.

[0003] Therefore, there is an urgent need for a high-precision and reliable method to predict protein-protein interactions that have not yet been revealed. With the continuous advancement of neural network technology, deep learning and graph neural networks have gradually been applied to the prediction of protein-protein interactions. PPI prediction methods based on neural networks can be divided into two categories: one is based only on amino acid sequences, and the other is based on the combination of amino acid sequences and protein structure information. The method that relies solely on amino acid sequences represents proteins by extracting sequence features. This method ignores the structural information of proteins, and therefore affects the accuracy of prediction to a certain extent. In contrast, the method that combines amino acid sequences and protein structural features not only considers sequence information, but also incorporates structural features, which can express the characteristics of proteins more comprehensively. This method makes up for the shortcomings of relying solely on sequences and enriches feature expression, but the extraction process of protein structural features is very complicated, which increases the difficulty of calculation and implementation of the model.

[0004] In summary, the laboratory method for predicting protein interactions is inefficient, time-consuming and laborious; the method that relies solely on amino acid sequences ignores the structural information of proteins; the method that combines amino acid sequences and protein structural features makes up for the lack of structural information, but the process of extracting structural features is very complicated. Therefore, a protein interaction prediction method based on a hybrid graph attention neural network is proposed. Summary of the invention

[0005] The technical problem to be solved by the present invention is: how to extract key features that can represent proteins, and based on these representative features, accurately and efficiently predict the interactions between proteins. At the same time, the prediction is realized through machine automation to reduce the cost of laboratory research, and a protein interaction prediction method based on a hybrid graph attention neural network is provided.

[0006] The present invention solves the above technical problems through the following technical solutions, and the present invention comprises the following steps:

[0007] S1: Preprocessing of protein data

[0008] Obtain multiple benchmark data sets, filter the benchmark data sets, and then merge the benchmark data sets to obtain a training set and a test set;

[0009] S2: Extract and fuse protein features

[0010] The evolutionary characteristics of the protein sequence are obtained through the amino acid sequence of the protein, and the atomic characteristics and secondary structure characteristics of the protein are obtained according to the protein PDB file. Then, the position embedding characteristics of the protein are calculated according to the position coordinates of the amino acid residues. Finally, the evolutionary characteristics, secondary structure characteristics, atomic characteristics, and position embedding characteristics of the protein are fused to obtain the characteristic expression of the protein as the input data feature of the protein-protein interaction prediction model.

[0011] S3: Building a TCN Neural Network

[0012] The encoding part of the transformer is spliced ​​and fused with the AGCN network to obtain the TCN network;

[0013] S4: Building a Hybrid Graph Attention Neural Network

[0014] Construct a hybrid graph attention neural network, fuse the GCN network with the GAT network to extract the local features of proteins, and use the TCN network to extract the global features of proteins, and obtain a protein-protein interaction hybrid graph attention neural network;

[0015] S5: Model training

[0016] The training set is used to train the protein-protein interaction mixed graph attention neural network to obtain the trained weight parameters, and the weight parameters are saved to obtain the protein-protein interaction prediction model;

[0017] S6: Protein-protein interaction prediction

[0018] The test set data is input into the protein-protein interaction prediction model for prediction to obtain the prediction results.

[0019] Furthermore, in the step S1, the specific processing process is as follows: obtaining the amino acid sequence of the protein and the public interaction data set, and performing six-step filtering processing on the data set in sequence, and the six-step filtering processing is to exclude structures with more than 30% missing residues, remove repeated chains with the same UniProtKB / Swiss-Prot identifier, remove transmembrane proteins, remove oligomeric structures, remove proteins with buried surface accessibility and interface polarity under a set threshold, and remove redundant proteins.

[0020] Furthermore, in step S2, the specific processing process is as follows:

[0021] S21: PSSM and HMM features are calculated through the amino acid sequence of the protein. After splicing the PSSM and HMM features, the evolutionary characteristics of the protein are obtained. The structure and sequence information of the protein are obtained through the PDB file of the protein, and the secondary structure feature DSSP of the protein is calculated according to the DSSP algorithm;

[0022] S22: According to the protein PDB file on the protein database website and the intrinsic characteristics of the residue, seven features of each atom constituting the residue are extracted: atomic mass, B factor, whether it is a side chain atom of the residue, electronic charge, the number of hydrogen atoms bonded to it, whether it is cyclic, and the van der Waals radius of the atom; the corresponding average values ​​of the seven features of all atoms constituting the residue are taken as the seven atomic features of the residue;

[0023] S23: The coordinates of the center of mass of the residue side chain are used as the residue pseudo-position, and the position embedding features of the protein are calculated using the Euclidean distance formula. The reference position for calculating the position embedding features is the first residue in each protein sequence.

[0024] S24: Finally, the four protein features of protein evolutionary features, secondary structure features, atomic features and position embedding features are fused and spliced ​​to obtain the 62-dimensional features of protein.

[0025] Furthermore, in step S3, the encoding part of the transformer is connected in series with the AGCN network to form a TCN network.

[0026] Furthermore, in step S4, the GCN network and the GAT network are connected in parallel to obtain the embedding representation of each node respectively.

[0027] Furthermore, in step S4, the calculation formula of the GCN network is as follows:

[0028] H (l+1) =σ(((1-α)PH (l) +αH(0) )((1-β l )I n +β l W (l) ))

[0029] Among them, σ represents the RELU activation function, P represents the normalized adjacency matrix in the equation, and I n represents the identity matrix, H (0) represents the input node feature matrix, H (l) and H (l+1) represents the node feature embedding before and after the layer is run, W (l) represents the weight matrix, α and β l is a hyperparameter.

[0030] Furthermore, in step S4, the calculation formula of the GAT network is as follows:

[0031] H (l+1) =σ((1-β l )((1-α)H (l) +αH (0) )+(β l (H (l) ||H (0) )W (l) ))

[0032] Among them, σ represents the RELU activation function, P represents the normalized adjacency matrix in the equation, and I n represents the identity matrix, H (0) represents the input node feature matrix, H (l) and H (l+1) represents the node feature embedding before and after the layer is run, || represents the connection operation, and W (l) Denotes linear learnable parameters, α and β l is a hyperparameter.

[0033] Furthermore, in step S4, in the protein-protein interaction mixed graph attention neural network, the feature expression of each protein is called a node, and the protein feature expression directly connected to the protein feature expression to be updated is called the neighborhood node of the protein.

[0034] Furthermore, in step S5, during the model training process, the loss function is a cross entropy loss function, and the formula is as follows:

[0035]

[0036] Where N is the total number of samples, y i is the true label of the i-th sample, p i is the predicted probability of the ith sample.

[0037] Furthermore, in step S6, the prediction results of protein-protein interactions are calculated with the original labels to obtain the performance parameters of the model.

[0038] Compared with the prior art, the present invention has the following advantages: the protein interaction prediction method based on the hybrid graph attention neural network fuses two features as protein feature expression, uses the TCN network to extract the global features of the protein, and does not need to extract the global features through the convolutional neural network; the sampling GAT and GCN networks extract the local features of the protein, and there is no need to extract the local features separately through the sliding window technology; both the GAT and GCN networks use residual structures and identity mappings, which can well alleviate the transition smoothing problem of deep GAT and GCN networks, and deeper protein features can be extracted through this operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is a schematic flow chart of a protein interaction prediction method based on a hybrid graph attention neural network in an embodiment of the present invention;

[0040] Figure 2 Schematic diagram of the TCN structure in the protein-protein interaction prediction model in an embodiment of the present invention;

[0041] Figure 3 Schematic diagram of the GCN and GAT structures in the protein-protein interaction prediction model in the embodiment of the present invention;

[0042] Figure 4 It is a schematic diagram of the overall structure of the protein-protein interaction prediction model in the example of the present invention. DETAILED DESCRIPTION

[0043] The following is a detailed description of an embodiment of the present invention. This embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation method and a specific operation process are given, but the protection scope of the present invention is not limited to the following embodiment.

[0044] This embodiment provides a technical solution: a protein interaction prediction method based on a hybrid graph attention neural network, comprising:

[0045] like Figure 1 As shown, this embodiment provides a technical solution: a protein interaction prediction method based on a hybrid graph attention neural network, comprising the following steps:

[0046] S1: Preprocessing of protein data.

[0047] In step S1, the following contents are specifically included:

[0048] S11: The protein data were refined through a six-step filtering process, including excluding structures with more than 30% missing residues, removing duplicate chains with the same UniProtKB / Swiss-Prot identifiers, removing transmembrane proteins, removing oligomeric structures (higher than dimers), removing proteins with buried surface accessibility and interface polarity below a certain threshold, and removing redundant proteins; the processed data and the distribution of interacting and non-interacting residues in the data are shown in Table 1:

[0049] Table 1 Statistics of three benchmark datasets and the training and test sets used in this example

[0050]

[0051]

[0052] S2: Extract multiple features of proteins and perform feature fusion;

[0053] In step S2, the following contents are specifically included:

[0054] S21: Obtain protein sequence and structural features through computational programs;

[0055] In step S21, the evolutionary characteristics of the protein are obtained through the amino acid sequence of the protein, including the position-specific score matrix (PSSM) and the hidden Markov model matrix (HMM). The PSSM is generated by running the alignment tool PSI-BLAST v2.10.1, which has an E value of 0.001 after three iterations. The HMM matrix is ​​generated by running the HHblits v3.0.3 algorithm using default parameters. DSSP is a program that calculates three types of structural characteristics of the protein. The structure and sequence information of the protein is obtained through the PDB file of the protein, and the secondary structure feature matrix (DSSP) of the protein is obtained according to the DSSP algorithm tool. Among them, the shapes of the PSSM and HMM matrices are both N v ×20, the DSSP matrix shape is N v ×14.

[0056] S22: Obtain the atomic features of the protein through the protein PDB file, and then calculate the positional embedding features of the protein through the positional coordinates of the amino acid residues;

[0057] In step S22, seven features of each atom (excluding hydrogen atoms) constituting the residue are extracted: atomic mass, B factor, whether it is a side chain atom of the residue, electronic charge, the number of hydrogen atoms bonded to it, whether it is cyclic, and the van der Waals radius of the atom; since the number of atoms of different residues is different, the corresponding average values ​​of the seven features of all atoms constituting the residue are taken as the seven atomic features of the residue, and the matrix shape is N v ×7; the position embedding feature of the residue contains the relative position information of each residue and the reference residue; in this embodiment, the coordinates of the center of mass of the residue side chain are used as the residue pseudo position, and the shape is N v ×3, the matrix shape is N calculated by the Euclidean distance formula v ×1 protein embedding features.

[0058] S22: Finally, the above four features are fused. PSSM and HMM are first spliced ​​to form the evolutionary characteristics of the protein, and then the secondary structure characteristics of the protein are spliced. Finally, the atomic characteristics and position embedding characteristics of the protein are spliced ​​to obtain the 62-dimensional characteristics of the protein as the input data features of the protein-protein interaction prediction model.

[0059] S3: Building TCN network

[0060] In step S3, the encoding part of the transformer is fused with the AGCN network to form a TCN network, which is used to extract the global features of proteins.

[0061] In step S3, the encoding part of the transformer is fused with the AGCN network. The fusion method is to obtain a TCN network by connecting the transformer encoding part and AGCN in series, which helps to improve the expression and generalization capabilities of the model.

[0062] S4: Building a Hybrid Graph Attention Neural Network

[0063] In step S4, a protein-protein interaction hybrid graph attention neural network is constructed with protein feature expressions as nodes and protein-protein interactions as edges. In this network, the GCN network and the GAT network add residual structure and identity mapping, which effectively alleviates the over-smoothing problem caused by deep graph neural networks, and extracts richer node features through two different models.

[0064] In this embodiment, in the protein-protein interaction hybrid graph attention neural network, the fine-grained relationship (local features) captured by the self-attention mechanism and the global structural features (global features) extracted by the TCN network are fused through the combination of the GAT network and the GCN network. This fusion not only enriches the feature space of the model, but also enhances the differences between nodes and improves the discriminability of classification and prediction tasks. At the same time, adding residuals to the deep graph neural network helps to alleviate the over-smoothing problem and enhance the discriminability of feature expression, and the residual connection facilitates the training of deep networks. The formula of the GCN network is as follows:

[0065] H (l+1) =σ(((1-α)PH (l) +αH (0) )((1-β l )I n +β l W (l) ))

[0066] Among them, σ represents the RELU activation function, P represents the normalized adjacency matrix in the equation, and I n represents the identity matrix, H (0) represents the input node feature matrix, H (l) and H (l+1) represents the node feature embedding before and after the layer is run, W (l) represents the weight matrix, α and β l is a hyperparameter.

[0067] The formula of the GAT network is as follows:

[0068] H (l+1) =σ((1-β l )((1-α)H (l) +αH (0) )+(β l (H (l) ||H (0) )W (l) ))

[0069] Among them, σ represents the RELU activation function, P represents the normalized adjacency matrix in the equation, and I n represents the identity matrix, H (0) represents the input node feature matrix, H (l) and H (l+1) represents the node feature embedding before and after the layer is run, || represents the connection operation, and W (l) Denotes linear learnable parameters, α and β l is a hyperparameter.

[0070] It should be noted that the above two formulas use initial residual and identity mapping calculations to alleviate the transition smoothing problem of the eight-layer GAT and GCN networks.

[0071] It should be noted that the residual connection adds the output of the previous layer to the output of the current layer as the final output of the layer, forming a "skip connection" to help information flow and avoid gradient disappearance. Identity mapping means that there is no change or convolution operation, and the input is directly passed to the output.

[0072] S5: Model network training

[0073] The training set is used to train the hybrid graph attention neural network to obtain a protein-protein interaction prediction model;

[0074] Specifically, in step S5, the specific training scheme is as follows:

[0075] S51: The initial learning rate is set to 0.001, and the model is optimized using the cross entropy loss function and the Adam optimizer. Each training includes 50 epochs.

[0076] S52: In this experiment, 5-fold cross-validation training was performed on the training set.

[0077] S53: During the training process, in order to demonstrate the relative importance of each feature we adopted, feature ablation experiments were performed by using a feature alone or excluding a feature from the final feature combination. The model performance of AUROC and AUPRC was evaluated using 5-fold cross validation (CV) and independent Test_60. Table 2 shows the impact of these features on model performance:

[0078] Table 2 Using individual features or removing individual features from the combined features in sequence

[0079]

[0080] S6: Protein-protein interaction prediction

[0081] The test set data (Test_60) is input into the protein-protein interaction prediction model for prediction, and the prediction results are obtained. The results are shown in Table 4. It can be seen from Table 4 that the performance of the present invention is better than other methods. The first five methods are based on sequences, and the last four methods use protein structure information. As shown in Table 3, our model (ours) is significantly better than other methods in seven indicators on the benchmark dataset Test_60.

[0082] Table 3 Performance comparison with other methods on Test_60

[0083]

[0084]

[0085] In summary, the protein interaction prediction method based on the hybrid graph attention neural network in the above embodiment first adopts three public and widely used benchmark data sets in previous studies: Dset_186, Dset_72 and Dset_164, which are named after the number of proteins in the data set. The above data set is refined through a six-step filtering process, including excluding structures with more than 30% missing residues, removing chains with the same UniprotKB / Swiss-Prot germplasm, removing transmembrane proteins, removing oligomeric structures (higher than dimers), removing proteins with buried surface accessibility and interface polarity under a certain threshold, and removing redundant proteins. A protein-protein interaction hybrid graph attention neural network is constructed with protein expression as nodes and protein-protein interactions as edges, and then the network model is trained. Here, the cross entropy loss function is used to accurately measure the binary classification of the data and predict the error, and it can also drive the model to optimize towards higher accuracy and confidence, thereby improving the model performance. Finally, the unknown protein is input into the protein-protein interaction prediction model for prediction to obtain the prediction result. By fusing the four features of proteins, constructing TCN networks, GAT networks and GCN networks to extract global and local features of proteins, the performance of different models was verified and compared on the test set. The experimental results show that these techniques can accelerate the convergence and robustness of the network; the model proposed in this invention has high prediction accuracy. In the future, it is expected to be deployed on edge devices for production and application.

[0086] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A protein interaction prediction method based on a hybrid graph attention neural network, characterized in that: The following steps are involved: S1: Preprocessing of protein data Obtain multiple benchmark data sets, filter the benchmark data sets, and then merge the benchmark data sets to obtain a training set and a test set; S2: Extract and fuse protein features The evolutionary characteristics of the protein sequence are obtained through the amino acid sequence of the protein, and the atomic characteristics and secondary structure characteristics of the protein are obtained according to the protein PDB file. Then, the position embedding characteristics of the protein are calculated according to the position coordinates of the amino acid residues. Finally, the evolutionary characteristics, secondary structure characteristics, atomic characteristics, and position embedding characteristics of the protein are fused to obtain the characteristic expression of the protein as the input data feature of the protein-protein interaction prediction model. S3: Building a TCN Neural Network The encoding part of the transformer is spliced ​​and fused with the AGCN network to obtain the TCN network; S4: Building a Hybrid Graph Attention Neural Network Construct a hybrid graph attention neural network, fuse the GCN network with the GAT network to extract the local features of proteins, and use the TCN network to extract the global features of proteins, and obtain a protein-protein interaction hybrid graph attention neural network; S5: Model training The training set is used to train the protein-protein interaction mixed graph attention neural network to obtain the trained weight parameters, and the weight parameters are saved to obtain the protein-protein interaction prediction model; S6: Protein-protein interaction prediction The test set data is input into the protein-protein interaction prediction model for prediction to obtain the prediction results.

2. The protein interaction prediction method based on hybrid graph attention neural network according to claim 1, characterized in that: In step S1, the specific processing process is as follows: obtaining the amino acid sequence of the protein and the public interaction data set, and performing six-step filtering processing on the data set in sequence, wherein the six-step filtering processing is to exclude structures with more than 30% missing residues, remove repeated chains with the same UniProtKB / Swiss-Prot identifier, remove transmembrane proteins, remove oligomeric structures, remove proteins with buried surface accessibility and interface polarity under a set threshold, and remove redundant proteins.

3. The protein interaction prediction method based on hybrid graph attention neural network according to claim 1, characterized in that: In step S2, the specific processing process is as follows: S21: PSSM and HMM features are calculated through the amino acid sequence of the protein. After splicing the PSSM and HMM features, the evolutionary characteristics of the protein are obtained. The structure and sequence information of the protein are obtained through the PDB file of the protein, and the secondary structure feature DSSP of the protein is calculated according to the DSSP algorithm; S22: According to the protein PDB file on the protein database website and the intrinsic characteristics of the residue, seven features of each atom constituting the residue are extracted: atomic mass, B factor, whether it is a side chain atom of the residue, electronic charge, the number of hydrogen atoms bonded to it, whether it is cyclic, and the van der Waals radius of the atom; the corresponding average values ​​of the seven features of all atoms constituting the residue are taken as the seven atomic features of the residue; S23: The coordinates of the center of mass of the residue side chain are used as the residue pseudo-position, and the position embedding features of the protein are calculated using the Euclidean distance formula. The reference position for calculating the position embedding features is the first residue in each protein sequence. S24: Finally, the four protein features of protein evolutionary features, secondary structure features, atomic features and position embedding features are fused and spliced ​​to obtain the 62-dimensional features of protein.

4. The protein interaction prediction method based on hybrid graph attention neural network according to claim 3, characterized in that: In step S3, the encoding part of the transformer is connected in series with the AGCN network to form a TCN network.

5. The protein interaction prediction method based on hybrid graph attention neural network according to claim 4, characterized in that: In step S4, the GCN network and the GAT network are connected in parallel to obtain the embedding representation of each node respectively.

6. The protein interaction prediction method based on hybrid graph attention neural network according to claim 5, characterized in that: In step S4, the calculation formula of the GCN network is as follows: H (l+1) =σ(((1-α)PH (l) +αH (0) )((1-b l )I n +b l W (l) )) Among them, σ represents the RELU activation function, P represents the normalized adjacency matrix in the equation, and I n represents the identity matrix, H (0) represents the input node feature matrix, H (l) and H (l+1) represents the node feature embedding before and after the layer is run, W (l) represents the weight matrix, α and β l is a hyperparameter.

7. The protein interaction prediction method based on hybrid graph attention neural network according to claim 5, characterized in that: In step S4, the calculation formula of the GAT network is as follows: H (l+1) =σ((1-β l )((1-a)H (l) +αH (0) )+(β l (H (l) ||H (0) )W (l) )) Among them, σ represents the RELU activation function, P represents the normalized adjacency matrix in the equation, and I n represents the identity matrix, H (0) represents the input node feature matrix, H (l) and H (l+1) represents the node feature embedding before and after the layer is run, || represents the connection operation, and W (l) Denotes linear learnable parameters, α and β l is a hyperparameter.

8. The protein interaction prediction method based on hybrid graph attention neural network according to claim 1, characterized in that: In step S4, in the protein-protein interaction mixed graph attention neural network, the feature expression of each protein is called a node, and the protein feature expression directly connected to the protein feature expression to be updated is called the neighborhood node of the protein.

9. The protein interaction prediction method based on hybrid graph attention neural network according to claim 1, characterized in that: In step S5, during the model training process, the loss function is a cross entropy loss function, and the formula is as follows: Where N is the total number of samples, y i is the true label of the i-th sample, p i is the predicted probability of the ith sample.

10. The protein interaction prediction method based on hybrid graph attention neural network according to claim 1, characterized in that: In step S6, the prediction results of protein-protein interactions are calculated with the original labels to obtain the performance parameters of the model.

Images