Protein docking method and apparatus, electronic device and storge medium
The iterative use of multiple molecular docking schemes and graph neural networks enhances protein docking accuracy, addressing low accuracy in existing methods and facilitating precise protein interaction predictions.
Patent Information
- Application Number
- JP2024143225
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-11-01
- Filing Date
- 2024-08-23
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-08-23
AI Technical Summary
Existing protein docking methods suffer from low docking accuracy.
A method involving iterative protein docking using at least two types of molecular docking schemes, such as rigid and flexible docking, with iteration termination conditions to generate a final complex conformation, and utilizing graph neural networks to enhance accuracy.
Improves the accuracy of protein docking by comprehensively considering multiple docking strategies and real-time updates, enabling precise prediction of protein interactions for applications like antigen-antibody complexes, polypeptide-protein complexes, and disease mechanism research.
Smart Images

Figure 2024167284000001_ABST
Abstract
Description
[Technical field]
[0001] The present disclosure relates to the field of artificial intelligence technology, in particular to the field of biocomputing, molecular docking, and deep learning technology, and in particular to a protein docking method, apparatus, electronic device, storage medium, and computer program. [Background technology]
[0002] Protein docking refers to the process of two or more protein molecules interacting with each other to form a complex, which plays a very important role in the biological field, such as antigen-antibody docking, polypeptide protein docking, etc. However, the protein docking method in the related art has a problem of low docking accuracy. Summary of the Invention
[0003] The present disclosure provides methods, apparatus, electronic devices, storage media and computer programs for protein docking.
[0004] According to a first aspect of the present disclosure, there is provided a protein docking method, the method including: in an iterative process of each round, docking a first protein and a second protein according to at least two types of molecular docking schemes to generate a complex conformation; and, if it is recognized that an iteration termination condition is not satisfied, continuing the iterative process of the next round until the iteration termination condition is satisfied to obtain a final complex conformation.
[0005] According to a second aspect of the present disclosure, there is provided a protein docking apparatus, the apparatus including: a docking module for docking a first protein and a second protein according to at least two types of molecular docking schemes in an iterative process of each round to generate a complex conformation; and a processing module for, when it is recognized that an iteration termination condition is not satisfied, continuing the iterative process of the next round until the iteration termination condition is satisfied to obtain a final complex conformation.
[0006] According to a third aspect of the present disclosure, there is provided an electronic device comprising at least one processor and a memory communicatively connected to said at least one processor, said memory storing instructions executable by said at least one processor, said instructions being executed by said at least one processor to cause said at least one processor to perform the protein docking method provided by the first aspect above.
[0007] According to a fourth aspect of the present disclosure, there is provided a non-transitory computer readable storage medium having stored thereon computer instructions, said computer instructions causing a computer to perform the protein docking method provided by the first aspect above.
[0008] According to a fifth aspect of the present disclosure, there is provided a computer program which, when executed by a processor, realises the protein docking method provided by the first aspect above.
[0009] It should be noted that the contents described in this section are not used to identify the essential or important features of the embodiments of the present disclosure, and are not intended to limit the scope of the present disclosure. Other features of the present disclosure will be more easily understood from the following description. [Brief description of the drawings]
[0010] The drawings are for a better understanding of the present solution and are not intended to limit the present disclosure. [Figure 1] 1 is a schematic flow chart of a protein docking method according to one embodiment of the present disclosure. [Diagram 2] 1 is a schematic flow chart of a protein docking method according to another embodiment of the present disclosure. [Diagram 3] 1 is a schematic flow chart of a protein docking method according to another embodiment of the present disclosure. [Figure 4] 1 is a schematic flow chart of a protein docking method according to another embodiment of the present disclosure. [Diagram 5] FIG. 1 is a schematic structural diagram of a protein docking apparatus according to one embodiment of the present disclosure. [Figure 6] FIG. 1 is a schematic block diagram of an electronic device according to one embodiment of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0011] In the following, exemplary embodiments of the present disclosure will be described in conjunction with the drawings, and various details of the embodiments of the present disclosure will be included for ease of understanding, and they should be regarded as illustrative. Therefore, as will be understood by those skilled in the art, various changes and modifications can be made to the embodiments without departing from the scope and spirit of the present disclosure. Similarly, for clarity and conciseness, the following description omits descriptions of well-known functions and structures.
[0012] AI (Artificial Intelligence) is a technical science that studies and develops theories, methods, technologies and application systems that simulate and extend human intelligence. At present, AI technology has the advantages of high automation, high accuracy and low cost, and is widely used.
[0013] Biocomputing refers to a computing model that treats biological macromolecules as "data" and is divided into three main types: protein computing, RNA (Ribonucleic Acid) computing, and DNA (Deoxyribo Nucleic Acid) computing. It also refers to a subfield of computer science and computer engineering that uses biotechnology and biology to build computers, while others, such as bioinformatics, are interdisciplinary sciences that use computers to store and process biological data.
[0014] Molecular docking is a method for designing drugs based on the characteristics of receptors and the interaction between receptors and drug molecules. It is mainly a theoretical simulation method for studying the interactions between molecules (such as ligands and receptors) and predicting their binding modes and affinities. In recent years, molecular docking has become an important technique in the field of computer-aided drug research.
[0015] DL (Deep Learning) is a new research direction in the field of ML (Machine Learning). It is a science that learns the inherent rules and expression levels of sample data, enabling machines to have human-like analytical learning abilities and recognize data such as text, images, and voice. It is widely applied to voice recognition and image recognition.
[0016] 1 is a schematic flow chart of a protein docking method according to an embodiment of the present disclosure. As shown in FIG. 1, the method includes the following steps S101 to S102.
[0017] In S101, in an iterative process of each round, a first protein and a second protein are docked according to at least two kinds of molecular docking methods to generate a complex conformation.
[0018] In addition, the execution entity of the protein docking method of the embodiment of the present disclosure may be a hardware device having data information processing capability and / or software required to drive the operation of the hardware device. Optionally, the execution entity may include a workstation, a server, a computer, a user terminal and other intelligent devices. Here, the user terminal may include, but is not limited to, a mobile phone, a computer, an intelligent voice interaction device, a smart home appliance, an in-vehicle terminal, etc.
[0019] It should be noted that the at least two kinds of molecular docking methods may include rigid docking, flexible docking, semi-flexible docking, etc., and are not excessively limited. Here, in the rigid docking process, neither the monomer conformation of the first protein nor the monomer conformation of the second protein changes. In the flexible docking process, at least one of the monomer conformations of the first protein and the monomer conformation of the second protein changes.
[0020] In addition, the execution order of the at least two kinds of molecular docking methods is not excessively limited. For example, in the case where the at least two kinds of molecular docking methods include rigid docking and flexible docking, the first protein and the second protein may be docked in the order of rigid docking and flexible docking to generate a complex conformation, or the first protein and the second protein may be docked in the order of flexible docking and rigid docking to generate a complex conformation.
[0021] In addition, each molecular docking method can generate a complex conformation. If a complex conformation does not currently exist, a complex conformation can be generated according to a molecular docking method. If a complex conformation currently exists, a complex conformation can be newly generated according to a molecular docking method, that is, the complex conformation can be updated.
[0022] For example, in the iterative process of each round, a first protein and a second protein are docked according to a rigid docking scheme to generate a complex conformation, and the complex conformation is updated according to a flexible docking scheme.
[0023] For example, in the iterative process of each round, a first protein and a second protein are docked according to a flexible docking scheme to generate a complex conformation, and the complex conformation is updated according to a rigid docking scheme.
[0024] If it is determined in S102 that the iteration termination condition is not met, the iteration process continues for the next round until the iteration termination condition is met to obtain the final complex conformation.
[0025] It should be noted that the iteration termination conditions are not overly restrictive. For example, they may include that the number of iteration rounds reaches a set threshold, that the parameters of the complex conformation are within a set range, that the first protein and the second protein are successfully docked, etc. Here, the parameters of the complex conformation are used to characterize the docking accuracy.
[0026] In one embodiment, obtaining the final complex conformation comprises taking the last obtained complex conformation as the final complex conformation.
[0027] In one embodiment, obtaining the final complex conformation comprises selecting the final complex conformation from a plurality of complex conformations generated by multiple rounds of the iterative process.
[0028] The protein docking method provided by the present disclosure is applicable to at least one of the following scenarios. Scene 1, Docking of antigen and antibody: The present disclosure can be used to predict antigen-antibody complex conformations and thus aid in antibody design. Scene 2, Polypeptide-protein docking: The present disclosure can be used to predict polypeptide-protein complex conformations, thereby aiding in the design of polypeptide pharmaceuticals. Scene 3, Disease Mechanism Research: The occurrence and progression of many diseases are related to abnormal interactions between proteins, and protein docking can help researchers understand the molecular mechanisms of these abnormal interactions, thus providing new ideas for disease diagnosis and treatment. Scene 4, Protein Function Research: Protein docking can reveal the interactions between proteins, which helps scientists understand the functions and control mechanisms of proteins. This is very important for elucidating biological processes such as cell signaling and gene regulation. Scene 5, Structural Biology: In protein structural biology, protein docking can be used to predict binding patterns in protein complexes, thus helping researchers explain the structure and function of protein complexes. Scene 6, Disease Mechanism Research: The occurrence and progression of many diseases are related to abnormal interactions between proteins. Protein docking can help researchers understand the molecular mechanisms of these abnormal interactions, thus providing new ideas for disease diagnosis and treatment. Scene 7, protein interaction network analysis: By predicting the interactions between proteins, a protein interaction network can be constructed, which helps researchers understand the interconnection and regulatory network of proteins in cells. Scene 8, Protein Engineering: In the biotechnology field, protein docking can be used to design new protein structures, such as fusion proteins, antibodies, enzymes, etc., to achieve specific functions.
[0029] In the protein docking method provided by the present disclosure, in each round of iteration process, a first protein and a second protein are docked according to at least two molecular docking methods to generate a complex conformation, and when it is recognized that the iteration end condition is not met, the next round of iteration process is continued until the iteration end condition is met to obtain a final complex conformation. In this way, multiple rounds of protein docking can be performed by comprehensively considering at least two molecular docking methods to realize protein docking, that is, multiple rounds of iteration of complex conformation are realized, and the accuracy of protein docking is improved.
[0030] Based on any one of the above embodiments, a target graph for generating a complex conformation can be constructed.
[0031] The construction of the target graph can be further understood in combination with FIG. 2, which is a schematic flowchart of a protein docking method in another embodiment of the present disclosure. As shown in FIG. 2, the method includes the following steps S201 to S203.
[0032] In S201, information on residues of the first protein is obtained based on the monomer conformation of the first protein.
[0033] In S202, information on residues of the second protein is obtained based on the monomer conformation of the second protein.
[0034] The information on the residue is not excessively limited, and may include, for example, the type of the residue and the position of the residue (e.g., the position of the α-carbon atom of the residue), where the α-carbon atom refers to the first carbon atom of the residue.
[0035] In addition, the monomer conformation of a protein carries information about the residues of the protein, and the information about the residues of the protein can be extracted from the monomer conformation of the protein.
[0036] For example, information on the residues of a first protein can be extracted from the monomer conformation of the first protein.
[0037] In S203, a target graph is constructed based on the residue information of the first protein and the residue information of the second protein, a node in the target graph represents a residue of the first protein or a residue of the second protein, an etch in the target graph represents an etch between two residues, and the target graph is used to generate a complex conformation.
[0038] Note that the target graph is a heterograph, i.e., the target graph contains two types of nodes, where the first type of nodes represent residues of a first protein and the second type of nodes represent residues of a second protein. The nodes in the target graph have a one-to-one correspondence with the residues. An etch in the target graph represents an etch between the residues corresponding to the two nodes connected by the etch.
[0039] The number of nodes in the target graph is the sum of the number of residues in the first protein and the number of residues in the second protein.
[0040] Note that there may or may not be an etch between any two nodes in the target graph.
[0041] In an embodiment of the present disclosure, constructing a target graph based on residue information of a first protein and residue information of a second protein includes several possible embodiments as follows.
[0042] Method 1: Generate a target graph by constructing nodes that have a one-to-one correspondence with residues, where the residues include residues of a first protein and residues of a second protein.
[0043] In this way, the method can realize the construction of nodes by constructing nodes that correspond one-to-one with residues.
[0044] Method 2: determine a distance between the feature of the first residue and the feature of the second residue based on the information of the first residue and the information of the second residue, where the first residue and the second residue belong to a target protein, the target protein is the first protein or the second protein, sort a plurality of second residues in ascending order according to the distance, and take the top N sorted second residues as target residues, where N is a positive integer, and generate a target graph by adding a connection etch between the node corresponding to the first residue and the node corresponding to the target residue.
[0045] Thus, the method determines the distance between the feature of the first residue and the feature of the second residue based on the information of the first residue and the information of the second residue, sets the N second residues with the closest distance as target residues, and adds a connection etch between the node corresponding to the first residue and the node corresponding to the target residue, thereby constructing etch between residues of the same protein.
[0046] It should be noted that the first and second residues belong to a target protein, i.e., the first and second residues belong to the same protein, e.g., the first and second residues belong to a first protein, or the first and second residues belong to a second protein.
[0047] It should be noted that N is not excessively limited and may be 10, for example.
[0048] In one embodiment, determining a distance between the first residue feature and the second residue feature based on the first residue information and the second residue information includes performing feature extraction on the first residue information to obtain the first residue feature, performing feature extraction on the second residue information to obtain the second residue feature, and obtaining a distance between the first residue feature and the second residue feature.
[0049] In one embodiment, the first residue information and the second residue information may be input into a KNN (K~Nearest Neighbor) model, and the target residue may be output by the KNN model.
[0050] The distance is not excessively limited, and may include, for example, Euclidean distance, Manhattan distance, and the like.
[0051] Method 3: Generate a target graph by taking the residue of the first protein as the third residue, taking the residue of the second protein as the fourth residue, and adding a connection etch between the node corresponding to the third residue and the node corresponding to the fourth residue.
[0052] This allows the method to construct etches between residues of different proteins by adding a connecting etch between a node corresponding to a third residue and a node corresponding to a fourth residue.
[0053] The third residue is any one of the residues in the first protein, and the fourth residue is any one of the residues in the second protein.
[0054] Method 4: Determine the features of a node based on the information of the residues corresponding to the node.
[0055] The features of the node are not excessively limited, and may include, for example, the type of residue corresponding to the node, the position of the residue corresponding to the node, and the like.
[0056] Method 5, determines the features of an Etch based on the information of the two residues corresponding to the Etch.
[0057] It should be noted that the characteristics of the etch are not overly limited and may include, for example, the length of the etch, the angle difference between two residues corresponding to the etch, and the like.
[0058] In one embodiment, determining the characteristic of the etch based on information of the two residues that correspond to the etch includes determining a distance between the two residues that correspond to the etch based on the positions of the two residues that correspond to the etch to be a length of the etch.
[0059] In one embodiment, determining the characteristic of the etch based on information of the two residues corresponding to the etch includes taking a difference between the angles of the two residues corresponding to the etch as the angle difference of the two residues corresponding to the etch.
[0060] Based on any one of the above embodiments, the method further includes: updating the residue information of the first protein and the residue information of the second protein according to the latest obtained complex conformation, and updating the target graph by returning to the step of constructing the target graph according to the residue information of the first protein and the residue information of the second protein, so that the residue information can be updated using the latest complex conformation, and the target graph can be reconstructed using the updated residue information, thereby realizing real-time updating of the target graph in the protein docking process.
[0061] In addition, the complex conformation carries information on the residues of the first protein and the residues of the second protein, and the information on the residues of the first protein and the residues of the second protein can be extracted from the most recently obtained complex conformation, and the original information on the residues of the first protein can be switched to the extracted information on the residues of the first protein, and the original information on the residues of the second protein can be switched to the extracted information on the residues of the second protein.
[0062] It should be noted that the update of the target graph may include, but is not limited to, removing or adding edges between nodes.
[0063] In some examples, after updating the residue information of the first protein and the residue information of the second protein, the method further includes updating a node feature and an etch feature in the target graph based on the residue information of the first protein and the residue information of the second protein, where the node feature is determined based on the residue information corresponding to the node, and the etch feature is determined by the information of two residues corresponding to the etch, so that the node feature and the etch feature in the target graph can be updated using the updated residue information, and real-time updating of the node feature and the etch feature can be realized in the docking process.
[0064] The protein docking method provided by the present disclosure can obtain residue information of a first protein based on the monomer conformation of the first protein, obtain residue information of a second protein based on the monomer conformation of the second protein, and construct a target graph based on the residue information of the first protein and the residue information of the second protein to generate a complex conformation.
[0065] In the above embodiment, the at least two types of molecular docking methods include rigid docking, and the step of docking a first protein and a second protein according to the at least two types of molecular docking methods in step S102 to generate a complex conformation may be further understood in combination with FIG. 3, which is a schematic flowchart of a protein docking method in another embodiment of the present disclosure. As shown in FIG. 3, the method includes the following S301 to S302.
[0066] In S301, a target graph is input to a first graph neural network, and a position of a keypoint in a complex conformation is obtained by the first graph neural network based on the target graph, and the keypoint is a position point on a contact surface between a first protein and a second protein in the complex conformation.
[0067] It should be noted that in the process of obtaining the positions of the key points in the complex conformation, neither the monomer conformation of the first protein nor the monomer conformation of the second protein changes.
[0068] In addition, the first graph neural network is not excessively limited, and may include, for example, a GCN (Graph Convolutional Network), a GRN (Graph Recurrent Network), a GAT (Graph Attention Network), etc.
[0069] It should be noted that the number of key points is not excessively limited.
[0070] In one embodiment, obtaining the positions of keypoints in the complex conformation based on the target graph by the first graph neural network includes updating node features and etch features in the target graph by the first graph neural network, and obtaining the positions of keypoints in the complex conformation based on the node features and etch features by the first graph neural network, thereby achieving obtaining the positions of keypoints in the complex conformation.
[0071] In addition, updating the node features and the edge features in the target graph by the first graph neural network can be realized by any one of the feature updating methods of graph neural networks in the related art, and is not excessively limited here.
[0072] In one embodiment, obtaining a position of the keypoint in the complex conformation based on the target graph by the first graph neural network includes obtaining positions of residues in the complex conformation based on the target graph by the first graph neural network, obtaining a distance between the third residue and the fourth residue based on a position of the third residue in the complex conformation and a position of the fourth residue in the complex conformation, and if the distance between the third residue and the fourth residue is smaller than a set threshold, determining a position in the complex conformation of a midpoint of the etch between the third residue and the fourth residue as the position of the keypoint in the complex conformation. In this example, if the distance between the third residue and the fourth residue is smaller than a set threshold, determining the midpoint of the etch between the third residue and the fourth residue as the keypoint.
[0073] In S302, a complex conformation is generated based on the positions of the key points in the complex conformation.
[0074] In one embodiment, generating a complex conformation based on the positions of the key points in the complex conformation includes determining a receptor and a ligand from the first protein and the second protein, obtaining a rotation translation matrix based on the positions of the key points in the monomer conformation of the ligand and the positions of the key points in the complex conformation, performing a global space transformation on the monomer conformation of the ligand based on the rotation translation matrix, and generating a complex conformation based on the transformed monomer conformation of the ligand and the monomer conformation of the receptor, whereby a rotation translation matrix can be obtained based on the positions of the key points in the complex conformation, performing a global space transformation on the monomer conformation of the ligand, and combining the transformed monomer conformation of the ligand and the monomer conformation of the receptor to generate a complex conformation.
[0075] It should be noted that performing a global spatial transformation on the monomer conformation of the ligand means performing a global rotation and a global translation on the monomer conformation of the ligand.
[0076] In some examples, determining a receptor and a ligand from the first protein and the second protein includes determining the first protein as a receptor and the second protein as a ligand, or determining the second protein as a receptor and the first protein as a ligand.
[0077] In some examples, obtaining a rotation translation matrix based on the positions of keypoints in the monomer conformation of the ligand and the positions of keypoints in the complex conformation includes: generating a first matrix based on the positions of a plurality of keypoints in the monomer conformation of the ligand; generating a second matrix based on the positions of a plurality of keypoints in the complex conformation, the second matrix being a product of the rotation translation matrix and the first matrix; and performing matrix decomposition on the second matrix based on the first matrix to obtain a rotation translation matrix. Note that the matrix decomposition can be realized by any one of matrix decomposition methods in the related art, and may include, for example, SVD (Singular Value Decomposition).
[0078] In some examples, obtaining a rotation translation matrix based on the positions of the keypoints in the monomer conformation of the ligand and the positions of the keypoints in the complex conformation includes setting elements in the rotation translation matrix as unknowns, constructing a system of equations based on the positions of the keypoints in the monomer conformation of the ligand, the positions of the keypoints in the complex conformation, and the elements in the rotation translation matrix, solving the system of equations, and obtaining the solutions of the system of equations as elements in the rotation translation matrix, thereby generating the rotation translation matrix.
[0079] In some examples, performing a global spatial transformation on the monomer conformation of the ligand based on the rotation translation matrix includes performing a spatial transformation on positions of residues of the ligand in the monomer conformation based on the rotation translation matrix to obtain a transformed monomer conformation of the ligand based on positions of a plurality of residues of the ligand in the monomer conformation.
[0080] In the protein docking method provided by the present disclosure, the at least two kinds of molecular docking methods include rigid docking, in which a target graph is input into a first graph neural network, and the first graph neural network obtains positions of key points in a complex conformation based on the target graph, the key points being position points at the contact surface between the first protein and the second protein in the complex conformation, and a complex conformation is generated based on the positions of the key points in the complex conformation, thereby realizing protein rigid docking.
[0081] In the above embodiment, the at least two types of molecular docking methods include flexible docking, and the step of docking a first protein and a second protein according to the at least two types of molecular docking methods in step S102 to generate a complex conformation may be further understood in combination with FIG. 4, which is a schematic flowchart of a protein docking method in another embodiment of the present disclosure. As shown in FIG. 4, the method includes the following steps S401 to S402.
[0082] In S401, the target graph is input into a second graph neural network, and the second graph neural network obtains residue positions in a complex conformation based on the target graph, and in the process of obtaining the residue positions in the complex conformation, at least one conformation of the monomer conformation of the first protein and the monomer conformation of the second protein is changed.
[0083] For the second graph neural network, please refer to the relevant description of the first graph neural network, and a detailed description will be omitted here.
[0084] The residues include residues of the first protein and residues of the second protein.
[0085] In one embodiment, obtaining residue positions in the complex conformation based on the target graph by the second graph neural network includes updating node features and etch features in the target graph by the second graph neural network, and obtaining residue positions in the complex conformation based on the node features and etch features by the second graph neural network, thereby achieving obtaining residue positions in the complex conformation.
[0086] In addition, updating the node features and the edge features in the target graph by the second graph neural network can be realized by any one of the feature updating methods of graph neural networks in the related art, and is not excessively limited here.
[0087] In S402, a complex conformation is generated based on the positions of multiple residues in the complex conformation.
[0088] In an embodiment of the present disclosure, generating a complex conformation based on positions of a plurality of residues in the complex conformation includes generating a complex conformation based on positions of residues of a plurality of first proteins in the complex conformation and positions of residues of a plurality of second proteins in the complex conformation.
[0089] In the protein docking method provided by the present disclosure, the at least two kinds of molecular docking methods include flexible docking, in which a target graph is input into a second graph neural network, and the second graph neural network obtains residue positions in a complex conformation based on the target graph, where in the process of obtaining the residue positions in the complex conformation, at least one of the monomer conformation of the first protein and the monomer conformation of the second protein is changed, and a complex conformation is generated based on the positions of multiple residues in the complex conformation, thereby realizing flexible docking of the protein.
[0090] Based on any one of the above embodiments, in an iterative process of each round, the first protein and the second protein may be docked according to the order of rigid docking and flexible docking to generate a complex conformation.
[0091] For example, a target graph can be constructed based on information on residues of a first protein and information on residues of a second protein, the target graph can be input to a first graph neural network, the positions of key points in a complex conformation can be obtained by the first graph neural network, and a complex conformation A can be generated based on the positions of the key points in the complex conformation.
[0092] Based on the complex conformation A, the residue information of the first protein and the residue information of the second protein may be updated, and the target graph may be updated by returning to the step of constructing a target graph based on the residue information of the first protein and the residue information of the second protein, and further, the node features and the etch features in the target graph may be updated based on the residue information of the first protein and the residue information of the second protein.
[0093] The target graph is input to a second graph neural network, and the second graph neural network obtains residue positions in a complex conformation based on the target graph, and generates a complex conformation B based on the positions of multiple residues in the complex conformation.
[0094] Based on any one of the above embodiments, in an iterative process of each round, the first protein and the second protein may be docked according to the order of flexible docking and rigid docking to generate a complex conformation.
[0095] For example, a target graph can be constructed based on information on residues of a first protein and information on residues of a second protein, the target graph can be input to a second graph neural network, and the positions of residues in a complex conformation can be obtained by the second graph neural network, and a complex conformation C can be generated based on the positions of multiple residues in the complex conformation.
[0096] Based on the complex conformation C, the residue information of the first protein and the residue information of the second protein may be updated, and the target graph may be updated by returning to the step of constructing a target graph based on the residue information of the first protein and the residue information of the second protein, and further the node features and etch features in the target graph may be updated based on the residue information of the first protein and the residue information of the second protein.
[0097] The target graph is input to a first graph neural network, and the first graph neural network obtains positions of key points in a complex conformation based on the target graph, and generates a complex conformation D based on the positions of key points in the complex conformation.
[0098] In the technical solution disclosed herein, the collection, storage, use, processing, transmission, provision, disclosure and other processing of relevant user personal information shall all comply with the provisions of relevant laws and regulations and shall not violate public order and morals.
[0099] According to an embodiment of the present disclosure, the present disclosure further provides a protein docking device for realizing the above protein docking method.
[0100] FIG. 5 is a block diagram of a protein docking apparatus according to one embodiment of the present disclosure.
[0101] As shown in FIG. 5, the protein docking apparatus 500 includes a docking module 501 and a processing module 502 .
[0102] The docking module 501 docks the first protein and the second protein according to at least two kinds of molecular docking schemes in an iterative process in each round to generate a complex conformation.
[0103] If the processing module 502 determines that the iteration termination condition is not met, it continues the iteration process for the next round until the iteration termination condition is met to obtain the final complex conformation.
[0104] In one embodiment of the present disclosure, the apparatus further includes a construction module, which obtains residue information of the first protein based on the monomer conformation of the first protein, obtains residue information of the second protein based on the monomer conformation of the second protein, and constructs a target graph based on the residue information of the first protein and the residue information of the second protein, where a node in the target graph represents a residue of the first protein or a residue of the second protein, an etch in the target graph represents an etch between two residues, and the target graph is used to generate the complex conformation.
[0105] In one embodiment of the present disclosure, the at least two kinds of molecular docking methods include rigid docking, and the docking module 501 further inputs the target graph into a first graph neural network, and obtains positions of key points in a complex conformation based on the target graph by the first graph neural network, where the key points are position points at the contact interface between the first protein and the second protein in the complex conformation, and generates the complex conformation based on the positions of the key points in the complex conformation.
[0106] In one embodiment of the present disclosure, the docking module 501 further updates node features and etch features in the target graph by the first graph neural network, and obtains the positions of the key points in a complex conformation based on the node features and the etch features by the first graph neural network.
[0107] In one embodiment of the present disclosure, the docking module 501 further determines a receptor and a ligand from the first protein and the second protein, obtains a rotation translation matrix based on the positions of the key points in the monomer conformation of the ligand and the positions of the key points in a complex conformation, performs a global space transformation on the monomer conformation of the ligand based on the rotation translation matrix, and generates the complex conformation based on the transformed monomer conformation of the ligand and the monomer conformation of the receptor.
[0108] In one embodiment of the present disclosure, the at least two kinds of molecular docking methods include flexible docking, and the docking module 501 further inputs the target graph into a second graph neural network, and obtains the residue positions in a complex conformation based on the target graph by the second graph neural network, where in the process of obtaining the residue positions in a complex conformation, at least one conformation among the monomer conformation of the first protein and the monomer conformation of the second protein is changed, and the complex conformation is generated based on the positions of multiple residues in the complex conformation.
[0109] In one embodiment of the present disclosure, the construction module further updates the residue information of the first protein and the residue information of the second protein based on a recently obtained complex conformation, and updates the target graph by returning to the step of constructing a target graph based on the residue information of the first protein and the residue information of the second protein.
[0110] In one embodiment of the present disclosure, after updating the residue information of the first protein and the residue information of the second protein, the construction module further updates a node feature and an etch feature in the target graph based on the residue information of the first protein and the residue information of the second protein, where the node feature is determined based on the residue information corresponding to the node, and the etch feature is determined by the information of two residues corresponding to the etch.
[0111] In one embodiment of the present disclosure, the construction module further determines a distance between the feature of the first residue and the feature of the second residue based on information of a first residue and information of a second residue, where the first residue and the second residue belong to a target protein, the target protein is the first protein or the second protein, sorting a plurality of second residues in ascending order based on the distance, and taking the sorted top N second residues as target residues, where N is a positive integer, and generating the target graph by adding a connection etch between the node corresponding to the first residue and the node corresponding to the target residue.
[0112] In one embodiment of the present disclosure, the construction module further generates the target graph by setting a residue of the first protein as a third residue, setting a residue of the second protein as a fourth residue, and adding a connection etch between a node corresponding to the third residue and a node corresponding to the fourth residue.
[0113] In the protein docking device provided by the present disclosure, in the iterative process of each round, a first protein and a second protein are docked according to at least two molecular docking methods to generate a complex conformation, and if it is recognized that the iteration end condition is not met, the iterative process of the next round is continued until the iteration end condition is met to obtain a final complex conformation. In this way, the protein docking can be realized by comprehensively considering at least two molecular docking methods, and multiple rounds of protein docking can be performed, that is, multiple rounds of iteration of the complex conformation can be realized, and the accuracy of protein docking can be improved.
[0114] According to an embodiment of the present disclosure, the present disclosure provides an electronic device, a readable storage medium, and a computer program.
[0115] 6 shows a schematic block diagram of an electronic device 600 that can be used to implement embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, mobile phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the practice of the present disclosure as described and / or claimed herein.
[0116] 6, the electronic device 600 includes a computing unit 601 that can perform various suitable operations and processes according to a computer program stored in a read-only memory (ROM) 602 or loaded from a storage unit 608 into a random access memory (RAM) 603. The RAM 603 may store various programs and data necessary for the operation of the electronic device 600. The computing unit 601, the ROM 602 and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.
[0117] Several components of the electronic device 600 are connected to an I / O interface 605, including input units 606 such as a keyboard, mouse, etc., output units 606 such as various types of displays, speakers, etc., storage units 608 such as magnetic disks, optical disks, etc., and communication units 609 such as a network card, modem, wireless communication transceiver, etc. The communication units 609 enable the electronic device 600 to exchange information / data with other equipment via computer networks such as the Internet and / or various telecommunication networks.
[0118] The computing unit 601 may be various general-purpose and / or special-purpose processing components having processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphic processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, computing units that execute various machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 executes each of the methods and processes described above, such as the protein docking method. For example, in some embodiments, the protein docking method may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 608. In some embodiments, some or all of the computer program may be loaded and / or installed in the electronic device 600 via the ROM 602 and / or the communication unit 609. When the computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps of the protein docking method described above may be performed. Alternatively, in other embodiments, the computing unit 601 may be configured to perform the protein docking method in any other suitable manner (eg, via firmware).
[0119] Various embodiments of the systems and techniques described herein may be realized in digital electronic circuitry systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include being embodied in one or more computer programs that may be executed and / or interpreted on a programmable system that includes at least one programmable processor, which may be a special purpose or general purpose programmable processor, and may include receiving data and instructions from and transmitting data and instructions to a storage system, at least one input device, and at least one output device.
[0120] The program codes for implementing the methods of the present disclosure can be programmed in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, a special purpose computer, or other programmable data processing apparatus, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are performed. The program codes may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.
[0121] In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain or store a program for use by or in combination with an instruction execution system, device, or apparatus. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium includes, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or apparatus, or any suitable combination of the above. More specific examples of machine-readable storage media include, but are not limited to, one or more wire-based electrical connections, portable computer disks, hard disks, random access memories (RAMs), read-only memories (ROMs), erasable programmable read-only memories (EPROMs or flash memories), optical fibers, portable compact disk read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the above.
[0122] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user, and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other types of devices can also provide interaction with a user, for example, feedback provided to the user can be any form of sensing feedback (e.g., visual feedback, auditory feedback, or haptic feedback), and input from the user can be received in any form (including sound input, speech input, or tactile input).
[0123] The systems and techniques described herein may be implemented in a computing system including a back-end component (e.g., a data server), or a computing system including a middleware component (e.g., an application server), or a computing system including a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user interacts with embodiments of the systems and techniques described herein), or in a computing system including any combination of such back-end, middleware, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communications network). Examples of communications networks include a local area network (LAN), a wide area network (WAN), and the Internet.
[0124] The computer system may include a client and a server. The client and server are generally remote from each other and typically interact through a communication network. The relationship between the client and the server is created by computer programs running on the corresponding computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.
[0125] According to an embodiment of the present disclosure, the present disclosure further provides a computer program, which, when executed by a processor, implements the steps of the protein docking method described in the above embodiment of the present disclosure.
[0126] It should be noted that, using the various forms of flows described above, steps can be rearranged, added, or deleted. For example, each step described in the present disclosure may be performed in parallel, sequentially, or in a different order, but this specification is not limited thereto as long as the desired results of the technical means disclosed in the present application can be achieved.
[0127] The above specific embodiments do not limit the scope of protection of the present disclosure. Those skilled in the art can make various modifications, combinations, subcombinations, and substitutions based on design requirements and other factors. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present disclosure should be included within the scope of protection of the present disclosure.
Claims
1. A method for docking a protein, comprising the steps of: docking a first protein and a second protein according to at least two molecular docking schemes in an iterative process of each round to generate a complex conformation; if it is determined that the iteration termination condition is not satisfied, continuing the iterative process for the next round until the iteration termination condition is satisfied to obtain a final complex conformation; A method for docking a protein comprising the steps of:
2. obtaining information on residues of the first protein based on the monomer conformation of the first protein; obtaining information on residues of the second protein based on the monomer conformation of the second protein; constructing a target graph based on the residue information of the first protein and the residue information of the second protein, where a node in the target graph represents a residue of the first protein or a residue of the second protein, and an etch in the target graph represents an etch between two residues, and the target graph is used to generate the complex conformation; The method for docking a protein according to claim 1 , further comprising:
3. The at least two molecular docking methods include rigid docking; The step of docking a first protein and a second protein according to at least two types of molecular docking methods to generate a complex conformation includes: inputting the target graph into a first graph neural network, and acquiring positions of key points in a complex conformation based on the target graph by the first graph neural network, the key points being position points at a contact surface between the first protein and the second protein in the complex conformation; generating a complex conformation based on the positions of the key points in the complex conformation; The method for docking a protein according to claim 2, comprising:
4. The step of obtaining positions of key points in a complex conformation based on the target graph by the first graph neural network includes: updating node characteristics and etch characteristics in the target graph with the first graph neural network; obtaining the locations of the key points in the complex conformation based on the node features and the etch features by the first graph neural network; The method for docking a protein according to claim 3, comprising:
5. generating the complex conformation based on the positions of the key points in the complex conformation, determining a receptor and a ligand from the first protein and the second protein; obtaining a rotation translation matrix based on the positions of the keypoints in the monomer conformation of the ligand and the positions of the keypoints in the complex conformation; performing a global spatial transformation on the monomer conformation of the ligand based on the rotation-translation matrix; generating the complex conformation based on the converted monomer conformation of the ligand and the monomer conformation of the receptor; The method for docking a protein according to claim 3, comprising:
6. The at least two molecular docking methods include flexible docking; The step of docking a first protein and a second protein according to at least two types of molecular docking methods to generate a complex conformation includes: inputting the target graph into a second graph neural network, and obtaining the positions of the residues in a complex conformation based on the target graph by the second graph neural network, wherein at least one of the conformation of the monomer of the first protein and the conformation of the monomer of the second protein is changed in the process of obtaining the positions of the residues in a complex conformation; generating a complex conformation based on positions of a plurality of residues in the complex conformation; The method for docking a protein according to claim 2, comprising:
7. updating information of residues of the first protein and information of residues of the second protein based on a recently obtained complex conformation; updating the target graph by returning to and performing the step of constructing a target graph based on the residue information of the first protein and the residue information of the second protein; The method for docking a protein according to claim 2, further comprising:
8. After updating the information on residues of the first protein and the information on residues of the second protein, The protein docking method of claim 7, further comprising a step of updating a node feature and an etch feature in the target graph based on information of residues of the first protein and information of residues of the second protein, wherein the node feature is determined based on information of a residue corresponding to the node, and the etch feature is determined based on information of two residues corresponding to the etch.
9. The step of constructing a target graph based on information of residues of the first protein and information of residues of the second protein includes: determining a distance between a feature of the first residue and a feature of the second residue based on information of a first residue and information of a second residue, wherein the first residue and the second residue belong to a target protein, and the target protein is the first protein or the second protein; sorting the second residues in ascending order according to the distance, and determining the top N sorted second residues as target residues, where N is a positive integer; generating the target graph by adding a connection etch between a node corresponding to the first residue and a node corresponding to the target residue; The method for docking a protein according to claim 2, comprising:
10. The step of constructing a target graph based on information of residues of the first protein and information of residues of the second protein includes: a residue of the first protein being a third residue; a residue of the second protein being a fourth residue; generating the target graph by adding a connection etch between a node corresponding to the third residue and a node corresponding to the fourth residue; The method for docking a protein according to claim 2, comprising:
11. A protein docking device comprising: a docking module for docking the first protein and the second protein according to at least two types of molecular docking schemes in an iterative process of each round to generate a complex conformation; a processing module for continuing the iterative process in the next round until the iterative termination condition is satisfied to obtain a final complex conformation if the iterative termination condition is recognized to be not satisfied; A protein docking device comprising:
12. Further comprising a construction module; The construction module comprises: obtaining information on residues of the first protein based on the monomer conformation of the first protein; obtaining information on residues of the second protein based on the monomer conformation of the second protein; The protein docking apparatus of claim 11, further comprising: constructing a target graph based on information of residues of the first protein and information of residues of the second protein; a node in the target graph represents a residue of the first protein or a residue of the second protein; an etch in the target graph represents an etch between two residues; and the target graph is used to generate the complex conformation.
13. The at least two molecular docking methods include rigid docking; The docking module further comprises: inputting the target graph into a first graph neural network; and acquiring positions of key points in a complex conformation based on the target graph by the first graph neural network, the key points being position points at a contact surface between the first protein and the second protein in the complex conformation; The protein docking apparatus according to claim 12 , which generates a complex conformation based on the positions of the key points in the complex conformation.
14. The docking module further comprises: updating node characteristics and etch characteristics in the target graph with the first graph neural network; The protein docking apparatus according to claim 13, wherein the first graph neural network obtains the positions of the key points in the complex conformation based on the node features and the etch features.
15. The docking module further comprises: determining a receptor and a ligand from the first protein and the second protein; obtaining a rotation translation matrix based on the positions of the keypoints in the monomer conformation of the ligand and the positions of the keypoints in the complex conformation; performing a global spatial transformation on the monomer conformation of the ligand based on the rotation / translation matrix; The protein docking apparatus according to claim 13, which generates the complex conformation based on the converted monomer conformation of the ligand and the monomer conformation of the receptor.
16. The at least two molecular docking methods include flexible docking; The docking module further comprises: inputting the target graph into a second graph neural network; obtaining positions of the residues in a complex conformation based on the target graph by the second graph neural network; and during the process of obtaining positions of the residues in a complex conformation, at least one of the monomer conformation of the first protein and the monomer conformation of the second protein is changed; The protein docking apparatus according to claim 12 , which generates a complex conformation based on the positions of multiple residues in the complex conformation.
17. The construction module further comprises: updating the residue information of the first protein and the residue information of the second protein based on the recently obtained complex conformation; The protein docking apparatus according to any one of claims 12 to 16, further comprising: updating the target graph by returning to and executing the step of constructing a target graph based on information on residues of the first protein and information on residues of the second protein.
18. After updating the residue information of the first protein and the residue information of the second protein, the construction module further comprises: The protein docking apparatus of claim 17, further comprising: updating node features and etch features in the target graph based on residue information of the first protein and residue information of the second protein, the node features being determined based on residue information corresponding to the node, and the etch features being determined based on information of two residues corresponding to the etch.
19. The construction module further comprises: determining a distance between a feature of the first residue and a feature of the second residue based on information of a first residue and information of a second residue, the first residue and the second residue belong to a target protein, the target protein being the first protein or the second protein; sorting the second residues in ascending order according to the distance, and determining the top N sorted second residues as target residues, where N is a positive integer; 17. The protein docking apparatus according to claim 12, further comprising: a step of generating the target graph by adding a connection etch between a node corresponding to the first residue and a node corresponding to the target residue.
20. The construction module further comprises: a residue of the first protein is a third residue; a residue of the second protein is a fourth residue; The protein docking apparatus according to any one of claims 12 to 16, further comprising: adding a connection etch between a node corresponding to the third residue and a node corresponding to the fourth residue to generate the target graph.
21. 1. An electronic device comprising: At least one processor; a memory communicatively coupled to the at least one processor; An electronic device, the memory storing instructions executable by the at least one processor, the instructions, when executed by the at least one processor, causing the at least one processor to perform the method of any one of claims 1 to 10.
22. A non-transitory computer-readable storage medium having computer instructions stored thereon, comprising: A non-transitory computer readable storage medium, the computer instructions causing a computer to perform the method of any one of claims 1 to 10.
23. A computer program which, when executed by a processor, implements the method of any one of claims 1 to 10.
Citation Information
Patent Citations
Method for calculating and simulating protein-protein docking
CN102314560A
Protein data processing method and device, electronic equipment and storage medium
CN116978450A
Unlocking de novo antibody design with generative artificial intelligence
WO2023154829A2
