Antibody design method and device, electronic equipment and storage medium

Abstract

The invention relates to an antibody design method and device, electronic equipment and a storage medium. Target antigen information and to-be-predicted antibody information correspondingly comprising at least one unknown amino acid are determined; and determining an antigen sequence representation matrix, an antibody sequence representation matrix and at least one feature pair representation matrix according to the target antigen information and the to-be-predicted antibody information, and further determining a corresponding candidate antibody atom structure. And under the condition that the candidate antibody atomic structure is incomplete, determining the candidate antibody atomic structure again according to the to-be-predicted antibody information after updating. And under the condition that the atomic structure of the candidate antibody is completed, determining that the atomic structure of the candidate antibody is the atomic structure of the target antibody matched with the target antigen. The antibody design can be automatically carried out according to the antigen and the to-be-predicted antibody, and the design process is simple and efficient. Meanwhile, antibody design can be carried out for many times in an iterative prediction mode, and an accurate design result is obtained by correcting the antibody structure in the circulation process.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to an antibody design method, device, electronic device and storage medium. Background Art

[0002] Protein design is a computational approach to developing new proteins that do not exist in nature. In one scenario, protein design is based on the original protein to improve its activity or increase its stability; in another scenario, protein design is based on generative AI methods to generate new proteins from scratch to achieve functions such as stabilizing motifs and binding to target proteins. At present, the accuracy and flexibility of related technologies in the scenario of generating proteins based on generative AI are not high. Summary of the invention

[0003] In view of this, the present disclosure proposes an antibody design method, device, electronic device and storage medium, aiming to improve the efficiency of the antibody design process, as well as the accuracy and flexibility of the results.

[0004] According to a first aspect of the present disclosure, a method for designing an antibody is provided, the method comprising:

[0005] Determining target antigen information and corresponding antibody information to be predicted, wherein the antibody to be predicted corresponding to the antibody information to be predicted includes at least one unknown amino acid;

[0006] Determine an antigen sequence representation matrix, an antibody sequence representation matrix, and at least one feature pair representation matrix according to the target antigen information and the antibody information to be predicted, wherein the antigen sequence representation matrix is ​​used to characterize homologous amino acid sequence features of the target antigen corresponding to the target antigen information, the antibody sequence representation matrix is ​​used to characterize amino acid sequence features of the antibody to be predicted, and the feature pair representation matrix is ​​used to characterize features between amino acid pairs;

[0007] Determining corresponding candidate antibody atomic structures according to the antigen sequence representation matrix, the antibody sequence representation matrix and the at least one feature pair representation matrix;

[0008] In response to the incomplete atomic structure of the candidate antibody, updating the information of the antibody to be predicted according to the atomic structure of the candidate antibody, and determining the atomic structure of the candidate antibody again according to the updated information of the antibody to be predicted;

[0009] In response to the candidate antibody atomic structure being complete, the candidate antibody atomic structure is determined to be the atomic structure of the target antibody that matches the target antigen.

[0010] In one possible implementation, the target antigen information includes the atomic structure of the target antigen, the amino acid sequence of the target antigen, and the antigenic epitope of the target antigen, and the antigenic epitope is used to characterize the amino acid position in the target antigen that interacts with the antibody to be predicted.

[0011] In a possible implementation manner, the information of the antibody to be predicted includes the amino acid sequence of the antibody to be predicted.

[0012] In a possible implementation, the determining of an antigen sequence representation matrix, an antibody sequence representation matrix, and at least one feature pair representation matrix according to the target antigen information and the antibody information to be predicted includes:

[0013] Determining an antigen sequence representation matrix by searching the amino acid sequence of the target antigen in a sequence database;

[0014] Determining an antibody sequence representation matrix based on the amino acid sequence of the antibody to be detected;

[0015] splicing the amino acid sequence of the target antigen and the amino acid sequence of the antibody to be detected to obtain a target amino acid sequence;

[0016] At least one feature representation matrix of the target amino acid sequence is determined according to the target antigen information and the antibody information to be detected.

[0017] In a possible implementation, determining the antigen sequence representation matrix by searching the amino acid sequence of the target antigen in a sequence database includes:

[0018] Searching the amino acid sequence of the target antigen in the sequence database to obtain a corresponding homologous amino acid sequence;

[0019] Clustering the homologous amino acid sequence and the amino acid sequence of the target antigen to obtain a preset number of similar amino acid sequences;

[0020] The probability distribution of the amino acid at each position in each type of amino acid sequence is calculated to obtain the corresponding antigen sequence representation matrix.

[0021] In a possible implementation, determining an antibody sequence representation matrix based on the amino acid sequence of the antibody to be detected includes:

[0022] Copying the amino acid sequence of the antibody to be detected multiple times to obtain a preset number of amino acid sequences of the antibody to be detected;

[0023] An antibody sequence representation matrix is ​​determined based on a preset number of replicated amino acid sequences of the antibodies to be detected.

[0024] In a possible implementation, the determining of at least one feature representation matrix of the target amino acid sequence according to the target antigen information and the antibody information to be detected includes:

[0025] According to the target antigen information and the antibody information to be detected, at least one of the target atom distance pair matrix, the amino acid type pair matrix, the mask type pair matrix, the chain type pair matrix, and the antigen epitope-antibody pair matrix corresponding to the target amino acid sequence is determined as a feature representation matrix, wherein the target atom distance pair matrix is ​​used to characterize the distance between target atoms between any two amino acids in the target amino acid sequence, the amino acid type pair matrix is ​​used to characterize the types of amino acids that may be included in any two amino acid positions in the target amino acid sequence, the mask type pair matrix is ​​used to characterize whether there are amino acids in any two amino acid positions in the target amino acid sequence, the chain type pair matrix is ​​used to characterize the chain type to which the amino acids in any two amino acid positions in the target amino acid sequence belong, and the antigen epitope-antibody pair matrix is ​​used to characterize whether each amino acid position between any two amino acid positions in the target amino acid sequence is an antigen epitope.

[0026] In a possible implementation, determining the corresponding candidate antibody atomic structure according to the antigen sequence representation matrix, the antibody sequence representation matrix and the at least one feature pair representation matrix includes:

[0027] splicing the antigen sequence representation matrix and the antibody sequence representation matrix to obtain a target sequence representation matrix;

[0028] The target sequence representation matrix and the at least one feature pair representation matrix are input into the trained structure prediction model, and the corresponding candidate antibody atomic structure is output.

[0029] In a possible implementation, updating the to-be-predicted antibody information according to the candidate antibody atomic structure, and re-determining the candidate antibody atomic structure according to the updated to-be-predicted antibody information, includes:

[0030] Determining a candidate antibody atomic sequence according to the candidate antibody atomic structure;

[0031] The candidate antibody atomic sequence is used as updated antibody information to be predicted corresponding to the antibody to be predicted, and the candidate antibody atomic structure is determined again according to the updated antibody information to be predicted.

[0032] In a possible implementation, determining the candidate antibody atomic sequence according to the candidate antibody atomic structure includes:

[0033] Determining corresponding structural feature information according to the candidate antibody atomic structure, wherein the structural feature information includes at least one of amino acid type, main chain characteristic angle, amino acid type pair, atomic pair distance and dihedral angle distance between every two amino acids;

[0034] The structural feature information is input into the trained sequence prediction network to determine the corresponding candidate antibody atomic sequence.

[0035] According to a second aspect of the present disclosure, there is provided an antibody design device, the device comprising:

[0036] An information determination module, used to determine target antigen information and corresponding antibody information to be predicted, wherein the antibody to be predicted corresponding to the antibody information to be predicted includes at least one unknown amino acid;

[0037] a matrix determination module, used to determine an antigen sequence representation matrix, an antibody sequence representation matrix and at least one feature pair representation matrix according to the target antigen information and the antibody information to be predicted, wherein the antigen sequence representation matrix is ​​used to characterize the homologous amino acid sequence characteristics of the target antigen corresponding to the target antigen information, the antibody sequence representation matrix is ​​used to characterize the amino acid sequence characteristics of the antibody to be predicted, and the feature pair representation matrix is ​​used to characterize the characteristics between amino acid pairs;

[0038] a structure determination module, configured to determine a corresponding candidate antibody atomic structure according to the antigen sequence representation matrix, the antibody sequence representation matrix and the at least one feature pair representation matrix;

[0039] a structure updating module, configured to update the antibody information to be predicted according to the candidate antibody atomic structure in response to the candidate antibody atomic structure being incomplete, and to determine the candidate antibody atomic structure again according to the updated antibody information to be predicted;

[0040] The antibody determination module is used to determine that the candidate antibody atomic structure is the atomic structure of the target antibody matching the target antigen corresponding to the target antigen information in response to the candidate antibody atomic structure being complete.

[0041] In one possible implementation, the target antigen information includes the atomic structure of the target antigen, the amino acid sequence of the target antigen, and the antigenic epitope of the target antigen, and the antigenic epitope is used to characterize the amino acid position in the target antigen that interacts with the antibody to be predicted.

[0042] In a possible implementation manner, the information of the antibody to be predicted includes the amino acid sequence of the antibody to be predicted.

[0043] In a possible implementation, the matrix determination module is further used to:

[0044] Determining an antigen sequence representation matrix by searching the amino acid sequence of the target antigen in a sequence database;

[0045] Determining an antibody sequence representation matrix based on the amino acid sequence of the antibody to be detected;

[0046] splicing the amino acid sequence of the target antigen and the amino acid sequence of the antibody to be detected to obtain a target amino acid sequence;

[0047] At least one feature representation matrix of the target amino acid sequence is determined according to the target antigen information and the antibody information to be detected.

[0048] In a possible implementation, the matrix determination module is further used to:

[0049] Searching the amino acid sequence of the target antigen in the sequence database to obtain a corresponding homologous amino acid sequence;

[0050] Clustering the homologous amino acid sequence and the amino acid sequence of the target antigen to obtain a preset number of similar amino acid sequences;

[0051] The probability distribution of the amino acid at each position in each type of amino acid sequence is calculated to obtain the corresponding antigen sequence representation matrix.

[0052] In a possible implementation, the matrix determination module is further used to:

[0053] Copying the amino acid sequence of the antibody to be detected multiple times to obtain a preset number of amino acid sequences of the antibody to be detected;

[0054] An antibody sequence representation matrix is ​​determined based on a preset number of replicated amino acid sequences of the antibodies to be detected.

[0055] In a possible implementation, the matrix determination module is further used to:

[0056] According to the target antigen information and the antibody information to be detected, at least one of the target atom distance pair matrix, the amino acid type pair matrix, the mask type pair matrix, the chain type pair matrix, and the antigen epitope-antibody pair matrix corresponding to the target amino acid sequence is determined as a feature representation matrix, wherein the target atom distance pair matrix is ​​used to characterize the distance between target atoms between any two amino acids in the target amino acid sequence, the amino acid type pair matrix is ​​used to characterize the types of amino acids that may be included in any two amino acid positions in the target amino acid sequence, the mask type pair matrix is ​​used to characterize whether there are amino acids in any two amino acid positions in the target amino acid sequence, the chain type pair matrix is ​​used to characterize the chain type to which the amino acids in any two amino acid positions in the target amino acid sequence belong, and the antigen epitope-antibody pair matrix is ​​used to characterize whether each amino acid position between any two amino acid positions in the target amino acid sequence is an antigen epitope.

[0057] In a possible implementation, the structure determination module is further used to:

[0058] splicing the antigen sequence representation matrix and the antibody sequence representation matrix to obtain a target sequence representation matrix;

[0059] The target sequence representation matrix and the at least one feature pair representation matrix are input into the trained structure prediction model, and the corresponding candidate antibody atomic structure is output.

[0060] In a possible implementation, the structure updating module is used to:

[0061] Determining a candidate antibody atomic sequence according to the candidate antibody atomic structure;

[0062] The candidate antibody atomic sequence is used as updated antibody information to be predicted corresponding to the antibody to be predicted, and the candidate antibody atomic structure is determined again according to the updated antibody information to be predicted.

[0063] In a possible implementation, the structure updating module is used to:

[0064] Determining corresponding structural feature information according to the candidate antibody atomic structure, wherein the structural feature information includes at least one of amino acid type, main chain characteristic angle, amino acid type pair, atomic pair distance and dihedral angle distance between every two amino acids;

[0065] The structural feature information is input into the trained sequence prediction network to determine the corresponding candidate antibody atomic sequence.

[0066] According to a third aspect of the present disclosure, an electronic device is provided, comprising: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to implement the above method when executing the instructions stored in the memory.

[0067] According to a fourth aspect of the present disclosure, a non-volatile computer-readable storage medium is provided, on which computer program instructions are stored, wherein the computer program instructions implement the above method when executed by a processor.

[0068] According to a fifth aspect of the present disclosure, a computer program product is provided, comprising a computer-readable code, or a non-volatile computer-readable storage medium carrying the computer-readable code. When the computer-readable code runs in a processor of an electronic device, the processor in the electronic device executes the above method.

[0069] In an embodiment of the present disclosure, the target antigen information and the corresponding antibody information to be predicted including at least one unknown amino acid are determined, and the antigen sequence representation matrix, the antibody sequence representation matrix and at least one feature pair representation matrix are determined based on the target antigen information and the antibody information to be predicted, and the corresponding candidate antibody atomic structure is further determined. In the case where the candidate antibody atomic structure is incomplete, the antibody information to be predicted and the candidate antibody atomic structure are determined again based on the updated antibody information to be predicted. When the candidate antibody atomic structure is complete, the candidate antibody atomic structure is determined to be the atomic structure of the target antibody that matches the target antigen. The present disclosure can automatically design antibodies based on antigens and antibodies to be predicted, and the design process is simple and efficient. At the same time, it is also possible to perform antibody design multiple times by iterative prediction, and correct the antibody structure during the cycle to obtain accurate design results.

[0070] Further features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the disclosure and, together with the description, serve to explain the principles of the disclosure.

[0072] Figure 1 A flow chart showing an antibody design method according to an embodiment of the present disclosure;

[0073] Figure 2 A schematic diagram showing a target antigen and a target antibody according to an embodiment of the present disclosure;

[0074] Figure 3 A schematic diagram showing a process of determining a target antibody according to an embodiment of the present disclosure;

[0075] Figure 4 A schematic diagram showing an antibody design device according to an embodiment of the present disclosure;

[0076] Figure 5 A schematic diagram showing an electronic device according to an embodiment of the present disclosure;

[0077] Figure 6 A schematic diagram showing another electronic device according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0078] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0079] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0080] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present disclosure.

[0081] The antibody design method of the disclosed embodiment can be performed by an electronic device such as a terminal device or a server. Among them, the terminal device can be any fixed or mobile terminal such as a user equipment (UE), a mobile device, a user terminal, a terminal, a cellular phone, a cordless phone, a personal digital assistant (PDA), a handheld device, a computing device, a vehicle-mounted device, a wearable device, etc. The server can be a separate server or a server cluster composed of multiple servers. Any electronic device can implement the antibody design method of the disclosed embodiment by calling a computer-readable instruction stored in a memory by a processor.

[0082] Figure 1 A flow chart of an antibody design method according to an embodiment of the present disclosure is shown. Figure 1 As shown, an antibody design method according to an embodiment of the present disclosure may include the following steps S10-S50.

[0083] Step S10: Determine the target antigen information and the corresponding antibody information to be predicted.

[0084] In one possible implementation, the electronic device first determines the target antigen for which the corresponding target antibody needs to be designed, and the antibody to be predicted for designing the target antibody, and determines the target antigen information of the target antigen and the antibody to be predicted information of the antibody to be predicted. Among them, the antibody to be predicted includes at least one unknown amino acid, and the disclosed embodiment predicts the type of unknown amino acid included therein to complete the corresponding amino acid sequence, thereby designing a target antibody that interacts with the target antigen. The target antigen information corresponding to the target antigen is used to characterize the amino acid characteristics of the target antigen, and the antibody to be predicted information corresponding to the antibody to be predicted is used to characterize the amino acid characteristics of the antibody to be predicted.

[0085] Optionally, the target antigen information may include the atomic structure of the target antigen, the amino acid sequence of the target antigen, and the antigenic epitope of the target antigen, where the antigenic epitope is used to characterize the amino acid position in the target antigen that interacts with the antibody to be predicted. The atomic structure of the target antigen can be characterized by a first structure matrix and a second structure matrix. The first structure matrix is The second structure matrix is The first structure matrix is ​​used to characterize the position of each atom in each amino acid in the target antigen, L1 is the length of the amino acid sequence of the target antigen, 37 is the dimension corresponding to the 37 atoms included in each amino acid, and 3 is the dimension corresponding to the three coordinates of each atom in the amino acid in the spatial coordinate system. The second structure matrix is ​​the atom mask matrix, which is used to characterize the type of atoms present in each amino acid. The amino acid sequence of the target antigen can be obtained by , where L1 is the length of the amino acid sequence of the target antigen, and the type of each amino acid in the amino acid sequence can be represented by an integer between 0 and 20, where 0-19 represents known amino acids and 20 represents unknown amino acids. Since the amino acid sequence of the target antigen is known, the amino acid sequence does not actually include an element with a value of 20, that is, it can be represented by Characterize the amino acid sequence of the target antigen. The epitope of the target antigen is used to characterize the amino acid position in the target antigen that interacts with the antibody to be detected. It can be pre-set and sequenced. Characterization, where L1 is the amino acid sequence length of the target antigen, when the corresponding amino acid position is the preset interaction position, the value at the position is determined to be 1, and when the corresponding amino acid position is not the preset interaction position, the value at the position is determined to be 0. The information of the antibody to be predicted may include the amino acid sequence of the antibody to be predicted, which can be obtained by sequence Characterization, where L2 is the length of the amino acid sequence of the antibody to be predicted, and the type of each amino acid in the amino acid sequence can be represented by an integer between 0 and 20, where 0-19 represents known amino acids and 20 represents unknown amino acids.

[0086] Figure 2Schematic diagram of a target antigen and a target antibody according to an embodiment of the present disclosure is shown. Figure 2 As shown, the atomic structure, amino acid sequence and antigen epitope of the target antigen are known target antigen information, and the amino acid sequence of the antibody to be detected corresponding to the target antigen, including known amino acids and unknown amino acids, is the antibody information to be detected. The electronic device can make predictions based on the target antigen information and the antibody information to be detected to obtain the atomic structure corresponding to the target antibody composed of the complete amino acid sequence.

[0087] Step S20: determining an antigen sequence representation matrix, an antibody sequence representation matrix and at least one feature pair representation matrix according to the target antigen information and the antibody information to be predicted.

[0088] In a possible implementation, after determining the target antigen information of the target antigen and the antibody information to be predicted of the antibody to be predicted, the electronic device performs feature extraction based on the current target antigen information and the antibody information to be predicted to obtain an antigen sequence representation matrix, an antibody sequence representation matrix, and at least one feature pair representation matrix. Among them, the antigen sequence representation matrix is ​​used to characterize the homologous amino acid sequence characteristics of the target antigen corresponding to the target antigen information, the antibody sequence representation matrix is ​​used to characterize the amino acid sequence characteristics of the antibody to be predicted, and the feature pair representation matrix is ​​used to characterize the characteristics between amino acid pairs.

[0089] Optionally, the electronic device can determine the antigen sequence representation matrix by searching the amino acid sequence of the target antigen in the sequence database, and determine the antibody sequence representation matrix based on the amino acid sequence of the antibody to be detected. The amino acid sequence of the target antigen and the amino acid sequence of the antibody to be detected are further spliced ​​to obtain the target amino acid sequence, and at least one feature representation matrix of the target amino acid sequence is determined based on the target antigen information and the antibody to be detected information. Among them, the antigen sequence representation matrix can be determined by first searching the amino acid sequence of the target antigen in the sequence database to obtain the corresponding N-1 homologous amino acid sequences, and then clustering the homologous amino acid sequences and the amino acid sequence of the target antigen, a total of N amino acid sequences, to obtain a preset number of amino acid sequence classes, and calculating the probability distribution of the amino acid at each position in each amino acid sequence class to obtain the corresponding antigen sequence representation matrix. The 21-dimensional parameter is used to characterize the probability that the amino acid in each amino acid position is of the 21 types of amino acids. The antigen sequence representation matrix can be used to limit the changes in the antigen structure and prevent the structure from being changed when the antigen interacts with the antibody.

[0090] Furthermore, since the antibody to be detected includes unknown amino acids, its corresponding homologous amino acid sequence cannot be searched from the sequence database, so the corresponding antibody sequence representation matrix can be directly determined based on the amino acid sequence of the antibody to be detected itself. That is, the electronic device can copy the amino acid sequence of the antibody to be detected multiple times to obtain a preset number of amino acid sequences of the antibody to be detected, and then determine the antibody sequence representation matrix based on the copied preset number of amino acid sequences of the antibody to be detected. Among them, the number of copies of the antibody to be detected can be the sum N of the amino acid sequence of the target antigen and the homologous amino acid sequence searched, that is, the amino acid sequences of N antibodies to be detected are copied to determine the antibody sequence representation matrix as The 21-dimensional parameters corresponding to each amino acid position in each amino acid sequence are one-hot sequences, which are used to characterize the amino acid type at that position.

[0091] In a possible implementation, at least one feature representation matrix of the target amino acid sequence determined based on the target antigen information and the antibody information to be detected may be determined based on the target antigen information and the antibody information to be detected. At least one of the target atom distance pair matrix, the amino acid type pair matrix, the mask type pair matrix, the chain type pair matrix, and the antigen epitope-antibody pair matrix corresponding to the target amino acid sequence is used as a feature representation matrix. Among them, the target atom distance pair matrix is ​​used to characterize the distance between target atoms between any two amino acids in the target amino acid sequence, the amino acid type pair matrix is ​​used to characterize the types of amino acids that may be included in any two amino acid positions in the target amino acid sequence, the mask type pair matrix is ​​used to characterize whether there are amino acids in any two amino acid positions in the target amino acid sequence, the chain type pair matrix is ​​used to characterize the chain type to which the amino acids in any two amino acid positions in the target amino acid sequence belong, and the antigen epitope-antibody pair matrix is ​​used to characterize whether each amino acid position between any two amino acid positions in the target amino acid sequence is an antigen epitope. Exemplarily, the target atom may be a Cα atom, that is, the dimension of the target atom distance pair matrix may be The 20-dimensional feature is used to characterize the distance between the Cα atoms of any two amino acids in the target amino acid sequence. The dimension of the amino acid type pair matrix can be The mask type can have a matrix dimension of The dimensions of the chain type pair matrix can be The dimension of the epitope-antibody pair matrix can be

[0092] Step S30: determining corresponding candidate antibody atomic structures according to the antigen sequence representation matrix, the antibody sequence representation matrix and the at least one feature pair representation matrix.

[0093] In a possible implementation, after the electronic device determines the antigen sequence representation matrix, the antibody sequence representation matrix and at least one feature pair representation matrix based on the target antigen information and the antibody information to be detected, the electronic device determines the corresponding candidate antibody atomic structure based on the above multiple representation matrices. Among them, the electronic device can first splice the antigen sequence representation matrix and the antibody sequence representation matrix to obtain the target sequence representation matrix. Then, the target sequence representation matrix and at least one feature pair representation matrix are input into the trained structure prediction model, and the structure prediction model predicts the atomic structure of the antibody to be predicted based on the target sequence representation matrix and at least one feature pair representation matrix, and outputs the corresponding candidate antibody atomic structure.

[0094] Optionally, the structure prediction model of the embodiment of the present disclosure can fill in the unknown amino acids in the antibody to be detected and output the corresponding atomic structure. In the case where there are a large number of unknown amino acids in the antibody to be detected, it may not be possible to fully fill in all the amino acids, and only some of the amino acids can be predicted and filled, that is, the candidate antibody atomic structure output by the structure prediction model can be a complete atomic structure or an incomplete atomic structure.

[0095] Furthermore, the structure prediction model of the disclosed embodiment draws on the Evoformer module and structure module of AlphaFold2. It takes the target sequence representation matrix and at least one feature pair representation matrix as input, and outputs a three-dimensional atomic model of the atomic structure of the antibody to be tested. The difference from AlphaFold2 is that some amino acids on the antibody are unknown, but the structure prediction model can still construct a three-dimensional atomic model of these amino acid skeletons. For residues with known amino acid types, atomic models of the side chains are also constructed.

[0096] Step S40: in response to the candidate antibody atomic structure being incomplete, updating the antibody information to be predicted according to the candidate antibody atomic structure, and determining the candidate antibody atomic structure again according to the updated antibody information to be predicted.

[0097] In a possible implementation, when the candidate antibody atomic structure predicted by the structural prediction model is not a complete atomic structure, the electronic device can update the antibody information to be predicted according to the candidate antibody atomic structure, so as to further redetermine the antigen sequence representation matrix, the antibody sequence representation matrix and at least one feature pair representation matrix based on the updated antibody information to be predicted. Further, the candidate antibody atomic structure is predicted based on the redetermined antigen sequence representation matrix, the antibody sequence representation matrix and at least one feature pair representation matrix by the structural prediction model to obtain an updated candidate antibody atomic structure.

[0098] Optionally, after obtaining an incomplete candidate antibody atomic structure, the electronic device may first determine the candidate antibody atomic sequence and then update the antibody information to be predicted based on the candidate antibody atomic sequence. That is, the candidate antibody atomic sequence may be first determined based on the candidate antibody atomic structure, and then the candidate antibody atomic sequence may be used as the updated antibody information to be predicted corresponding to the antibody to be predicted, and the candidate antibody atomic structure may be determined again based on the updated antibody information to be predicted. Among them, the candidate antibody atomic sequence may be predicted based on the trained sequence prediction network. Exemplarily, after obtaining the candidate antibody atomic structure, the electronic device may determine the corresponding structural feature information based on the candidate antibody atomic structure, the structural feature information including at least one of the amino acid type, main chain feature angle, amino acid type pair, atomic pair distance and dihedral angle distance between each two amino acids, and then input the structural feature information into the trained sequence prediction network to determine the corresponding candidate antibody atomic sequence. Among them, the Cα atom of each amino acid in the candidate antibody atomic structure is a node V in the graph. i , it is connected to the 25 nearest Cα atoms (j∈N{V i There is an edge E between ij The amino acid type and main chain characteristic angle are the characteristics of the node, and the amino acid type pair, atom pair distance and dihedral angle distance between every two amino acids are the characteristics of the edge formed by two nodes. The amino acid type indicates the type of amino acid corresponding to each atom. The main chain characteristic angle can include the values ​​of φ, ψ, and θ angles on the main chain. The amino acid type pair is used to characterize the types of two amino acids in adjacent positions. The atom pair distance is used to characterize the distance between the atom pairs of every two amino acids. The dihedral angles between every two amino acids include NC α -C β -C β , C α -C β -C β -C α , C β -C β -C α -Atomic dihedral angles of C.

[0099] Optionally, the sequence prediction network in the embodiment of the present disclosure includes an encoder and a decoder, wherein the network framework of the encoder and the decoder may be the same, including and E ij ←LayerNorm(E ij +MLP(V i , V j , E ij )) are used to process node features and edge features respectively. In the prediction process of the sequence prediction network, only the node features are updated. Finally, the type distribution of 20 amino acids corresponding to each node is predicted and sampled: Where T is the sampling temperature, the higher the temperature, the higher the sampling diversity.

[0100] Furthermore, after the electronic device predicts the candidate antibody atomic sequence through the sequence prediction network obtained through training, the candidate antibody atomic sequence is used as the updated antibody information to be predicted corresponding to the antibody to be predicted, and the antigen sequence representation matrix, the antibody sequence representation matrix and at least one feature pair representation matrix are re-determined based on the updated antibody information to be predicted. Further, the atomic structure of the candidate antibody is predicted based on the re-determined antigen sequence representation matrix, the antibody sequence representation matrix and at least one feature pair representation matrix through the structural prediction model to obtain an updated candidate antibody atomic structure. As the number of times the candidate antibody atomic structure is predicted cyclically increases, the sampling temperature T of the candidate antibody atomic sequence gradually decreases, making the result of sequence sampling more and more stable.

[0101] In a possible implementation, the structure prediction model and sequence prediction network of the embodiment of the present disclosure can be pre-trained and then retrained together. The process of retraining is to first obtain the three-dimensional structure files of two interacting proteins from the PDB database, and then randomly select a line from the corresponding amino acid sequence to perform random masking on the interaction interface as training data to jointly train the structure prediction model and sequence prediction network.

[0102] Step S50: In response to the candidate antibody atomic structure being complete, determining that the candidate antibody atomic structure is the atomic structure of the target antibody that matches the target antigen.

[0103] In one possible implementation, after the electronic device predicts the atomic structure of the candidate antibody at least once and obtains the complete atomic structure of the candidate antibody including all known amino acids, it can directly determine that the atomic structure of the candidate antibody is the atomic structure of the designed target antibody, that is, the atomic structure of the candidate antibody is an atomic structure that matches the target antigen and can interact with it.

[0104] Figure 3 A schematic diagram of a process for determining a target antibody according to an embodiment of the present disclosure is shown. Figure 3As shown, in the case of designing a target antibody that interacts with a target antigen, the disclosed embodiment can first determine the target antigen information and the corresponding antibody information to be predicted including at least one unknown amino acid, determine the antigen sequence representation matrix, the antibody sequence representation matrix and at least one feature pair representation matrix based on the target antigen information and the antibody information to be predicted, and further determine the corresponding candidate antibody atomic structure. In the case where the candidate antibody atomic structure is incomplete, the antibody information to be predicted and the candidate antibody atomic structure are determined again based on the updated antibody information to be predicted. In the case where the candidate antibody atomic structure is complete, the candidate antibody atomic structure is determined to be the atomic structure of the target antibody that matches the target antigen.

[0105] Based on the above technical features, the disclosed embodiment uses the current best protein structure prediction model AlphaFold2 as the basic module of the structure prediction model, which improves the accuracy of predicting the atomic structure of the antibody. At the same time, during the training process, the structure prediction model and the sequence prediction network are jointly trained specifically for the interaction interface to further improve the accuracy of the final designed antibody. Furthermore, since the sequence sampling is not fixed during multiple cycles of prediction, it can be corrected during multiple cycles of prediction to ensure the accuracy of the results. During use, the user only needs to provide partial sequences of the antigen structure and the antibody to achieve antibody design and antigen-antibody complex structure prediction. The entire prediction process is simple and efficient.

[0106] Figure 4 A schematic diagram of an antibody design device according to an embodiment of the present disclosure is shown. Figure 4 As shown, the antibody design device of the embodiment of the present disclosure may include:

[0107] An information determination module 40 is used to determine target antigen information and corresponding antibody information to be predicted, wherein the antibody to be predicted corresponding to the antibody information to be predicted includes at least one unknown amino acid;

[0108] A matrix determination module 41 is used to determine an antigen sequence representation matrix, an antibody sequence representation matrix and at least one feature pair representation matrix according to the target antigen information and the antibody information to be predicted, wherein the antigen sequence representation matrix is ​​used to characterize the homologous amino acid sequence characteristics of the target antigen corresponding to the target antigen information, the antibody sequence representation matrix is ​​used to characterize the amino acid sequence characteristics of the antibody to be predicted, and the feature pair representation matrix is ​​used to characterize the characteristics between amino acid pairs;

[0109] A structure determination module 42, configured to determine a corresponding candidate antibody atomic structure according to the antigen sequence representation matrix, the antibody sequence representation matrix and the at least one feature pair representation matrix;

[0110] a structure updating module 43, configured to update the to-be-predicted antibody information according to the candidate antibody atomic structure in response to the candidate antibody atomic structure being incomplete, and to re-determine the candidate antibody atomic structure according to the updated to-be-predicted antibody information;

[0111] The antibody determination module 44 is configured to determine, in response to the candidate antibody atomic structure being complete, that the candidate antibody atomic structure is an atomic structure of a target antibody that matches the target antigen corresponding to the target antigen information.

[0112] In one possible implementation, the target antigen information includes the atomic structure of the target antigen, the amino acid sequence of the target antigen, and the antigenic epitope of the target antigen, and the antigenic epitope is used to characterize the amino acid position in the target antigen that interacts with the antibody to be predicted.

[0113] In a possible implementation manner, the information of the antibody to be predicted includes the amino acid sequence of the antibody to be predicted.

[0114] In a possible implementation, the matrix determination module 41 is further configured to:

[0115] Determining an antigen sequence representation matrix by searching the amino acid sequence of the target antigen in a sequence database;

[0116] Determining an antibody sequence representation matrix based on the amino acid sequence of the antibody to be detected;

[0117] splicing the amino acid sequence of the target antigen and the amino acid sequence of the antibody to be detected to obtain a target amino acid sequence;

[0118] At least one feature representation matrix of the target amino acid sequence is determined according to the target antigen information and the antibody information to be detected.

[0119] In a possible implementation, the matrix determination module 41 is further configured to:

[0120] Searching the amino acid sequence of the target antigen in the sequence database to obtain a corresponding homologous amino acid sequence;

[0121] Clustering the homologous amino acid sequence and the amino acid sequence of the target antigen to obtain a preset number of similar amino acid sequences;

[0122] The probability distribution of the amino acid at each position in each type of amino acid sequence is calculated to obtain the corresponding antigen sequence representation matrix.

[0123] In a possible implementation, the matrix determination module 41 is further configured to:

[0124] Copying the amino acid sequence of the antibody to be detected multiple times to obtain a preset number of amino acid sequences of the antibody to be detected;

[0125] An antibody sequence representation matrix is ​​determined based on a preset number of replicated amino acid sequences of the antibodies to be detected.

[0126] In a possible implementation, the matrix determination module 41 is further configured to:

[0127] According to the target antigen information and the antibody information to be detected, at least one of the target atom distance pair matrix, the amino acid type pair matrix, the mask type pair matrix, the chain type pair matrix, and the antigen epitope-antibody pair matrix corresponding to the target amino acid sequence is determined as a feature representation matrix, wherein the target atom distance pair matrix is ​​used to characterize the distance between target atoms between any two amino acids in the target amino acid sequence, the amino acid type pair matrix is ​​used to characterize the types of amino acids that may be included in any two amino acid positions in the target amino acid sequence, the mask type pair matrix is ​​used to characterize whether there are amino acids in any two amino acid positions in the target amino acid sequence, the chain type pair matrix is ​​used to characterize the chain type to which the amino acids in any two amino acid positions in the target amino acid sequence belong, and the antigen epitope-antibody pair matrix is ​​used to characterize whether each amino acid position between any two amino acid positions in the target amino acid sequence is an antigen epitope.

[0128] In a possible implementation, the structure determination module 42 is further configured to:

[0129] splicing the antigen sequence representation matrix and the antibody sequence representation matrix to obtain a target sequence representation matrix;

[0130] The target sequence representation matrix and the at least one feature pair representation matrix are input into the trained structure prediction model, and the corresponding candidate antibody atomic structure is output.

[0131] In a possible implementation, the structure updating module 43 is used to:

[0132] Determining a candidate antibody atomic sequence according to the candidate antibody atomic structure;

[0133] The candidate antibody atomic sequence is used as updated antibody information to be predicted corresponding to the antibody to be predicted, and the candidate antibody atomic structure is determined again according to the updated antibody information to be predicted.

[0134] In a possible implementation, the structure updating module 43 is used to:

[0135] Determining corresponding structural feature information according to the candidate antibody atomic structure, wherein the structural feature information includes at least one of amino acid type, main chain characteristic angle, amino acid type pair, atomic pair distance and dihedral angle distance between every two amino acids;

[0136] The structural feature information is input into the trained sequence prediction network to determine the corresponding candidate antibody atomic sequence.

[0137] In some embodiments, the functions or modules included in the device provided by the embodiments of the present disclosure can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0138] The embodiment of the present disclosure also provides a computer-readable storage medium on which computer program instructions are stored, and the computer program instructions implement the above method when executed by a processor. The computer-readable storage medium can be a volatile or non-volatile computer-readable storage medium.

[0139] An embodiment of the present disclosure further proposes an electronic device, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to implement the above method when executing the instructions stored in the memory.

[0140] The embodiments of the present disclosure also provide a computer program product, including a computer-readable code, or a non-volatile computer-readable storage medium carrying the computer-readable code. When the computer-readable code runs in a processor of an electronic device, the processor in the electronic device executes the above method.

[0141] Figure 5 A schematic diagram of an electronic device 800 according to an embodiment of the present disclosure is shown. For example, the electronic device 800 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, etc.

[0142] Reference Figure 5 , the electronic device 800 may include one or more of the following components: a processing component 802 , a memory 804 , a power component 806 , a multimedia component 808 , an audio component 810 , an input / output (I / O) interface 812 , a sensor component 814 , and a communication component 816 .

[0143] The processing component 802 generally controls the overall operation of the electronic device 800, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the above-mentioned method. In addition, the processing component 802 may include one or more modules to facilitate the interaction between the processing component 802 and other components. For example, the processing component 802 may include a multimedia module to facilitate the interaction between the multimedia component 808 and the processing component 802.

[0144] The memory 804 is configured to store various types of data to support operations on the electronic device 800. Examples of such data include instructions for any application or method operating on the electronic device 800, contact data, phone book data, messages, pictures, videos, etc. The memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0145] The power supply component 806 provides power to the various components of the electronic device 800. The power supply component 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the electronic device 800.

[0146] The multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 808 includes a front camera and / or a rear camera. When the electronic device 800 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

[0147] The audio component 810 is configured to output and / or input audio signals. For example, the audio component 810 includes a microphone (MIC), and when the electronic device 800 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 804 or sent via the communication component 816. In some embodiments, the audio component 810 also includes a speaker for outputting audio signals.

[0148] I / O interface 812 provides an interface between processing component 802 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: home button, volume button, start button, and lock button.

[0149] The sensor assembly 814 includes one or more sensors for providing various aspects of status assessment for the electronic device 800. For example, the sensor assembly 814 can detect the open / closed state of the electronic device 800, the relative positioning of the components, such as the display and keypad of the electronic device 800, and the sensor assembly 814 can also detect the position change of the electronic device 800 or a component of the electronic device 800, the presence or absence of contact between the user and the electronic device 800, the orientation or acceleration / deceleration of the electronic device 800, and the temperature change of the electronic device 800. The sensor assembly 814 may include a proximity sensor configured to detect the presence of a nearby object without any physical contact. The sensor assembly 814 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0150] The communication component 816 is configured to facilitate wired or wireless communication between the electronic device 800 and other devices. The electronic device 800 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 816 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0151] In an exemplary embodiment, the electronic device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above methods.

[0152] In an exemplary embodiment, a non-volatile computer-readable storage medium is also provided, such as a memory 804 including computer program instructions, which can be executed by a processor 820 of an electronic device 800 to perform the above method.

[0153] Figure 6 1 is a schematic diagram showing another electronic device 1900 according to an embodiment of the present disclosure. For example, the electronic device 1900 may be provided as a server or a terminal device. Figure 6 , the electronic device 1900 includes a processing component 1922, which further includes one or more processors, and a memory resource represented by a memory 1932 for storing instructions executable by the processing component 1922, such as an application. The application stored in the memory 1932 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 1922 is configured to execute instructions to perform the above method.

[0154] The electronic device 1900 may also include a power supply component 1926 configured to perform power management of the electronic device 1900, a wired or wireless network interface 1950 configured to connect the electronic device 1900 to a network, and an input / output (I / O) interface 1958. The electronic device 1900 may operate based on an operating system stored in the memory 1932, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™ or the like.

[0155] In an exemplary embodiment, a non-volatile computer-readable storage medium is also provided, such as a memory 1932 including computer program instructions, which can be executed by the processing component 1922 of the electronic device 1900 to perform the above method.

[0156] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0157] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0158] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0159] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.

[0160] Various aspects of the present disclosure are described herein with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer-readable program instructions.

[0161] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processor of the computer or other programmable data processing device, a device that implements the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and / or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0162] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0163] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and the part of the module, program segment or instruction contains one or more executable instructions for realizing the specified logical function. In some alternative implementations, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or action, or can be implemented with a combination of special hardware and computer instructions.

[0164] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for designing an antibody, characterized in that: The method comprises: Determining target antigen information and corresponding antibody information to be predicted, wherein the antibody to be predicted corresponding to the antibody information to be predicted includes at least one unknown amino acid; Determine an antigen sequence representation matrix, an antibody sequence representation matrix, and at least one feature pair representation matrix according to the target antigen information and the antibody information to be predicted, wherein the antigen sequence representation matrix is ​​used to characterize homologous amino acid sequence features of the target antigen corresponding to the target antigen information, the antibody sequence representation matrix is ​​used to characterize amino acid sequence features of the antibody to be predicted, and the feature pair representation matrix is ​​used to characterize features between amino acid pairs; Determining corresponding candidate antibody atomic structures according to the antigen sequence representation matrix, the antibody sequence representation matrix and the at least one feature pair representation matrix; In response to the incomplete atomic structure of the candidate antibody, updating the information of the antibody to be predicted according to the atomic structure of the candidate antibody, and determining the atomic structure of the candidate antibody again according to the updated information of the antibody to be predicted; In response to the candidate antibody atomic structure being complete, the candidate antibody atomic structure is determined to be the atomic structure of the target antibody that matches the target antigen.

2. The method according to claim 1, characterized in that The target antigen information includes the atomic structure of the target antigen, the amino acid sequence of the target antigen, and the antigenic epitope of the target antigen, wherein the antigenic epitope is used to characterize the amino acid position in the target antigen that interacts with the antibody to be predicted.

3. The method according to claim 2, characterized in that The information of the antibody to be predicted includes the amino acid sequence of the antibody to be predicted.

4. The method according to claim 3, characterized in that: The step of determining an antigen sequence representation matrix, an antibody sequence representation matrix, and at least one feature pair representation matrix according to the target antigen information and the antibody information to be predicted comprises: Determining an antigen sequence representation matrix by searching the amino acid sequence of the target antigen in a sequence database; Determining an antibody sequence representation matrix based on the amino acid sequence of the antibody to be detected; splicing the amino acid sequence of the target antigen and the amino acid sequence of the antibody to be detected to obtain a target amino acid sequence; At least one feature representation matrix of the target amino acid sequence is determined according to the target antigen information and the antibody information to be detected.

5. The method according to claim 4, characterized in that The step of determining the antigen sequence representation matrix by searching the amino acid sequence of the target antigen in a sequence database comprises: Searching the amino acid sequence of the target antigen in the sequence database to obtain a corresponding homologous amino acid sequence; Clustering the homologous amino acid sequence and the amino acid sequence of the target antigen to obtain a preset number of similar amino acid sequences; The probability distribution of the amino acid at each position in each type of amino acid sequence is calculated to obtain the corresponding antigen sequence representation matrix.

6. The method according to claim 4 or 5, characterized in that: The step of determining an antibody sequence representation matrix based on the amino acid sequence of the antibody to be detected comprises: Copying the amino acid sequence of the antibody to be detected multiple times to obtain a preset number of amino acid sequences of the antibody to be detected; An antibody sequence representation matrix is ​​determined based on a preset number of replicated amino acid sequences of the antibodies to be detected.

7. The method according to any one of claims 4 to 6, characterized in that: The step of determining at least one characteristic representation matrix of the target amino acid sequence according to the target antigen information and the antibody information to be detected comprises: According to the target antigen information and the antibody information to be detected, at least one of the target atom distance pair matrix, the amino acid type pair matrix, the mask type pair matrix, the chain type pair matrix, and the antigen epitope-antibody pair matrix corresponding to the target amino acid sequence is determined as a feature representation matrix, wherein the target atom distance pair matrix is ​​used to characterize the distance between target atoms between any two amino acids in the target amino acid sequence, the amino acid type pair matrix is ​​used to characterize the types of amino acids that may be included in any two amino acid positions in the target amino acid sequence, the mask type pair matrix is ​​used to characterize whether there are amino acids in any two amino acid positions in the target amino acid sequence, the chain type pair matrix is ​​used to characterize the chain type to which the amino acids in any two amino acid positions in the target amino acid sequence belong, and the antigen epitope-antibody pair matrix is ​​used to characterize whether each amino acid position between any two amino acid positions in the target amino acid sequence is an antigen epitope.

8. The method according to any one of claims 1 to 7, characterized in that: The determining of the corresponding candidate antibody atomic structure according to the antigen sequence representation matrix, the antibody sequence representation matrix and the at least one feature pair representation matrix comprises: splicing the antigen sequence representation matrix and the antibody sequence representation matrix to obtain a target sequence representation matrix; The target sequence representation matrix and the at least one feature pair representation matrix are input into the trained structure prediction model, and the corresponding candidate antibody atomic structure is output.

9. The method according to any one of claims 1 to 8, characterized in that: The step of updating the antibody information to be predicted according to the candidate antibody atomic structure, and re-determining the candidate antibody atomic structure according to the updated antibody information to be predicted includes: Determining a candidate antibody atomic sequence according to the candidate antibody atomic structure; The candidate antibody atomic sequence is used as updated antibody information to be predicted corresponding to the antibody to be predicted, and the candidate antibody atomic structure is determined again according to the updated antibody information to be predicted.

10. The method according to claim 9, characterized in that Determining the candidate antibody atomic sequence according to the candidate antibody atomic structure comprises: Determining corresponding structural feature information according to the candidate antibody atomic structure, wherein the structural feature information includes at least one of amino acid type, main chain characteristic angle, amino acid type pair, atomic pair distance and dihedral angle distance between every two amino acids; The structural feature information is input into the trained sequence prediction network to determine the corresponding candidate antibody atomic sequence.

11. An antibody design device, characterized in that: The device comprises: An information determination module, used to determine target antigen information and corresponding antibody information to be predicted, wherein the antibody to be predicted corresponding to the antibody information to be predicted includes at least one unknown amino acid; a matrix determination module, used to determine an antigen sequence representation matrix, an antibody sequence representation matrix and at least one feature pair representation matrix according to the target antigen information and the antibody information to be predicted, wherein the antigen sequence representation matrix is ​​used to characterize the homologous amino acid sequence characteristics of the target antigen corresponding to the target antigen information, the antibody sequence representation matrix is ​​used to characterize the amino acid sequence characteristics of the antibody to be predicted, and the feature pair representation matrix is ​​used to characterize the characteristics between amino acid pairs; a structure determination module, configured to determine a corresponding candidate antibody atomic structure according to the antigen sequence representation matrix, the antibody sequence representation matrix and the at least one feature pair representation matrix; a structure updating module, configured to update the antibody information to be predicted according to the candidate antibody atomic structure in response to the candidate antibody atomic structure being incomplete, and to determine the candidate antibody atomic structure again according to the updated antibody information to be predicted; The antibody determination module is used to determine that the candidate antibody atomic structure is the atomic structure of the target antibody that matches the target antigen corresponding to the target antigen information in response to the candidate antibody atomic structure being complete.

12. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to implement the method described in any one of claims 1 to 10 when executing the instructions stored in the memory.

13. A non-volatile computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 10 is implemented.

Images