Antibody de novo design method and apparatus, device and storage medium

Abstract

Provided are an antibody de novo design method and apparatus, a device and a storage medium. The method comprises: acquiring the structure and an epitope residue sequence of a target antigen; on the basis of the structure and the epitope residue sequence of the target antigen, generating a library of related antibodies that can potentially bind to the target antigen epitope; on the basis of the structure and the epitope residue sequence of the target antigen as well as the structures and sequences of related antibodies in the library of related antibodies, predicting the affinity of the related antibodies to the target antigen epitope, and screening out alternative antibodies; on the basis of the degree of accuracy of a preset model, selecting a corresponding pretreatment method to treat the alternative antibodies, so as to obtain experimental antibody molecules and sequences; and verifying whether the experimental antibody molecules can bind to the target antigen epitope residue, and screening out, as target antibodies, experimental antibody molecules that can bind to the target antigen epitope residue. The provided solution can realize the antibody design for specified antigen epitopes without animal immunization.

Description

Antibody de novo design method, device, equipment and storage medium

[0001] This application claims priority to the Chinese patent application filed with the State Intellectual Property Office on September 26, 2023, with application number 2023112467042 and application name “Antibody de novo design method, device and equipment and storage medium”, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The present application relates to the fields of artificial intelligence technology and computational structural biology technology, and in particular to methods, devices, equipment, and storage media for de novo antibody design. Background Art

[0003] Antibodies, a type of immunoglobulin, are composed of variable and constant regions. The structure of the variable region determines binding specificity. Upon detecting an antigen (such as a foreign pathogen), the human immune system produces antibodies that specifically bind to the antigen, achieving precision therapy by lysing cells or directing phagocytes to engulf the specific antigen.

[0004] Traditional antibody design methods involve immunizing animals and isolating the resulting antibodies. However, due to limitations such as insufficient immunogenicity of the antigen and epitope itself, or the instability and difficulty of expressing the antigen protein itself, antibodies cannot be obtained through animal immunization, or antibodies targeting specific antigen epitopes cannot be obtained.

[0005] Summary of the Invention

[0006] In order to solve or partially solve the problems existing in the related art, the present application provides a method, device, equipment and storage medium for de novo antibody design, which can achieve antibody design against specified antigen epitopes without animal immunization.

[0007] The antibody de novo design method of the present application comprises:

[0008] Obtain the structure and epitope residue sequence of the target antigen;

[0009] generating a library of related antibodies that can potentially bind to the target antigen epitope based on the structure and epitope residue sequence of the target antigen, wherein the library comprises the structure and sequence of related antibodies;

[0010] According to the structure and epitope residue sequence of the target antigen, as well as the structure and sequence of the relevant antibodies in the relevant antibody library, the affinity of the relevant antibodies to the target antigen epitope is predicted based on a preset model, and candidate antibodies that meet the preset affinity conditions are screened;

[0011] According to the accuracy of the preset model, a corresponding preprocessing method is selected to process the candidate antibody to obtain an experimental antibody molecule and sequence;

[0012] Verify whether the experimental antibody molecule can bind to the target antigen epitope residue, and screen out the experimental antibody molecule that can bind to the target antigen epitope residue as the target antibody.

[0013] In one embodiment, the method of predicting the affinity of the relevant antibodies for the target antigen epitope based on a preset model according to the structure and epitope residue sequence of the target antigen, and the structure and sequence of the relevant antibodies in the relevant antibody library, and screening out candidate antibodies that meet the preset affinity conditions, comprises:

[0014] According to the structure and epitope residue sequence of the target antigen, as well as the structure and sequence of related antibodies in the related antibody library, affinity prediction is performed on the sequence of the related antibody with the epitope residue sequence of the target antigen, and on the structure of the related antibody with the structure of the target antigen based on a preset model to obtain the binding probability of the related antibody to the target antigen epitope;

[0015] From the related antibodies whose binding probability is a preset probability, antibodies with a preset proportion are screened out as candidate antibodies.

[0016] In one embodiment, screening out antibodies with a preset proportion from the relevant antibodies with a preset binding probability as candidate antibodies comprises:

[0017] From the sequences with the highest or second highest binding probability, 20%-30% of the sequences are selected as candidate antibodies.

[0018] In one embodiment, the candidate antibodies are subjected to humanization screening to obtain humanized antibodies and form a humanized antibody library.

[0019] In one embodiment, the humanization screening of the candidate antibodies to obtain humanized antibodies and form a humanized antibody library comprises:

[0020] Performing humanization screening on the candidate antibodies, eliminating antibodies composed of non-human parts that are immunogenic, and retaining antibodies composed of human parts that are non-immunogenic;

[0021] The non-immunogenic antibodies composed of human parts are regarded as humanized antibodies, and a humanized antibody library is formed.

[0022] In one embodiment, the method of selecting a corresponding pre-processing method to process the candidate antibody based on the accuracy of the preset model to obtain the experimental antibody molecule and sequence includes:

[0023] If the accuracy of the preset model is high, a first preprocessing method is used to process a first preset number of candidate antibodies to obtain experimental antibody molecules and sequences;

[0024] If the accuracy of the preset model is low, a second preprocessing method is used to process the humanized antibodies in the humanized antibody library, and a second preset number of experimental antibody molecules and sequences are screened from the processed humanized antibodies.

[0025] In one embodiment, the first pretreatment method comprises recombinant expression.

[0026] In one embodiment, the second pretreatment method comprises:

[0027] Dividing the humanized antibody into multiple fragments, connecting the fragments by molecular biological methods, and constructing the complete antibody sequence into a vector to form a library for display screening;

[0028] The complete antibody sequence is screened in the library to obtain the processed humanized antibody.

[0029] If the affinity of the candidate antibody does not reach the preset threshold, the candidate antibody is subjected to affinity maturation treatment to obtain a mature antibody with an affinity reaching the preset threshold.

[0030] In one embodiment, it is verified whether the epitope to which the experimental antibody molecule binds to the target antigen is the target epitope, and a target functional antibody that can bind to the target epitope of the target antigen is screened.

[0031] The antibody de novo design device of the present application comprises:

[0032] An acquisition module is used to obtain the structure and epitope residue sequence of the target antigen;

[0033] An antibody generation module, configured to generate a library of related antibodies that can potentially bind to the target antigen epitope based on the structure and epitope residue sequence of the target antigen, wherein the library of related antibodies comprises the structure and sequence of related antibodies;

[0034] An affinity prediction module, configured to predict the affinity of the relevant antibodies to the target antigen epitope based on a preset model according to the structure and epitope residue sequence of the target antigen, as well as the structure and sequence of the relevant antibodies in the relevant antibody library, and to screen out candidate antibodies that meet the preset affinity conditions;

[0035] A preprocessing module is used to select a corresponding preprocessing method to process the candidate antibody according to the accuracy of the preset model to obtain an experimental antibody molecule and sequence. The experimental antibody molecule can be used as an experimental subject. The experiment is used to verify whether the experimental antibody molecule can bind to the target antigen epitope residue, and to screen out the experimental antibody molecule that can bind to the target antigen epitope residue as the target antibody.

[0036] In one embodiment, the affinity prediction module includes a prediction module and a screening module. The prediction module is used to predict the affinity of the sequence of the relevant antibody with the epitope residue sequence of the target antigen, and the structure of the relevant antibody with the structure of the target antigen based on a preset model according to the structure and epitope residue sequence of the target antigen, as well as the structure of the relevant antibody with the structure of the target antigen, to obtain the binding probability of the relevant antibody with the target antigen epitope; the screening module is used to screen out antibodies with a preset proportion as alternative antibodies from the relevant antibodies with the preset binding probability.

[0037] In one embodiment, the device further comprises a humanization screening module, which is used to perform humanization screening on the candidate antibodies to obtain humanized antibodies and form a humanized antibody library.

[0038] The electronic device of the present application includes a processor and a memory, wherein the memory stores executable code. When the executable code is executed by the processor, the processor executes the above method.

[0039] The computer-readable storage medium of the present application stores executable code, which, when executed by a processor of an electronic device, causes the processor to perform the above-described method. Additional aspects and advantages of the embodiments of the present application will be described in part in the following description and will become apparent from the following description or learned through practice of the embodiments of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The above and other objects, features and advantages of the present application will become more apparent through a more detailed description of exemplary embodiments of the present application in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments of the present application.

[0041] FIG1 is a schematic flow chart of the method for de novo antibody design shown in an embodiment of the present application;

[0042] FIG2 is another schematic flow chart of the method for de novo antibody design shown in an embodiment of the present application;

[0043] FIG3 is a framework diagram of the de novo antibody design method shown in the Examples of the present application;

[0044] FIG4 is a schematic structural diagram of an antibody de novo design apparatus according to an embodiment of the present application;

[0045] FIG5 is a schematic structural diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0046] The following describes embodiments of the present application in more detail with reference to the accompanying drawings. Although the accompanying drawings illustrate embodiments of the present application, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. Rather, these embodiments are provided to make the present application more thorough and complete, and to fully convey the scope of the present application to those skilled in the art.

[0047] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. As used in this application and the appended claims, the singular forms "a," "an," "the," and "the" are intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0048] The present application is described in detail below. Examples of the present application are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0049] Please refer to Figure 1. This application provides a method for de novo antibody design, comprising:

[0050] 01: Obtain the structure and epitope residue sequence of the target antigen.

[0051] 02: Based on the structure and epitope residue sequence of the target antigen, a library of related antibodies that can potentially bind to the target antigen epitope is generated. The related antibody library contains the structure and sequence of related antibodies.

[0052] 03: Based on the structure and epitope residue sequence of the target antigen, as well as the structure and sequence of related antibodies in the relevant antibody library, the affinity of the relevant antibodies to the target antigen epitope is predicted based on the preset model, and candidate antibodies that meet the preset affinity conditions are screened.

[0053] 04: Based on the accuracy of the preset model, select the corresponding preprocessing method to process the candidate antibodies to obtain the experimental antibody molecules and sequences.

[0054] 05: Verify whether the experimental antibody molecule can bind to the target antigen epitope residue, and screen out the experimental antibody molecule that can bind to the target antigen epitope residue as the target antibody.

[0055] In the present application, the target antigen can be a protein, and the structure of the target antigen can be a structure predicted by experimental means (such as cryo-electron microscopy, X-ray diffraction) or computational methods (such as AlphaFold, etc.).

[0056] In this application, there is no limitation on the storage format of the target antigen structure. It can be stored in pdb format or cif format. For example, the structure of the antigen is 1ZVM.pdb, and the antigen epitope residue sequence can be the epitope residue sequence on any chain or multiple chains.

[0057] In the present application, the sequence of the target antigen is obtained to the epitope residue level, providing a basis for subsequently obtaining antibodies against the target antigen epitope.

[0058] In this application, the number of related antibodies in the related antibody library is in the tens of millions, for example, 10 million levels. The related antibodies are related to the target antigen epitope, so a large number of related antibodies can be generated, so that the related antibodies have diversity, and the diversity is mainly in the CDR region, so that there are more related antibodies in the related antibody library that can potentially bind to the target antigen epitope. After generating a large number of related antibody sequences, data support is provided for the next step of narrowing down the scope to antibodies that bind to a specific epitope.

[0059] The candidate antibodies of the present application are antibodies that have a high probability of binding to the target antigen epitope. Step S3 can screen out candidate antibodies in the following manner: based on the structure and epitope residue sequence of the target antigen, as well as the structure and sequence of related antibodies in the related antibody library, the sequence of the related antibody is compared with the epitope residue sequence of the target antigen, and the structure of the related antibody is compared with the structure of the target antigen based on a preset model to predict the affinity, and obtain the binding probability of the related antibody to the target antigen epitope; from the related antibodies with a preset binding probability, antibodies with a preset proportion are screened as candidate antibodies.

[0060] The prediction model in this application can be an antigen epitope-antibody affinity prediction model. The number of candidate antibodies screened is in the millions, for example, 2 million.

[0061] The present application can verify whether the experimental antibody molecule can bind to the target antigen epitope residue through the following steps: if the accuracy of the preset model is high, a first pretreatment method is used to process a first preset number of candidate antibodies to obtain experimental antibody molecules and sequences; if the accuracy of the preset model is low, a second pretreatment method is used to process the humanized antibodies in the humanized antibody library, and a second preset number of experimental antibody molecules and sequences are screened from the treated humanized antibodies.

[0062] The present application can verify whether the experimental antibody molecule can block the binding of known antibodies / ligands to target antigen proteins by means of wet experiments such as ELISA / SPR / BLI, to verify whether the binding epitope of the experimental antibody molecule to the target antigen meets the design prediction, and the experimental antibody molecule that can block the binding of known antibodies / ligands to the target antigen protein is used as the target antibody. The complex structure of the experimental antibody molecule and the target antigen can also be obtained by the method of cryo-electron microscopy to verify whether the binding epitope of the experimental antibody molecule to the target antigen meets the design prediction, and the experimental antibody molecule that meets the design prediction is used as the target antibody. And the cell function experiment related to the antigen function can also be used to detect whether the experimental antibody molecule has the expected biological function, and the disease animal model related to the antigen function can also be used to detect whether the experimental antibody molecule can treat the corresponding disease in the animal model.

[0063] The present application can also verify whether the epitope bound to the target antigen by the experimental antibody molecule is correct through methods such as CryoEM and HDX, thereby finding the target functional antibody that can bind to the target epitope, which is more helpful in finding and confirming the antibody containing the target function, and can obtain dozens of target antibody sequences that bind to the correct epitope.

[0064] In summary, this application targets a user-specified target antigen epitope, designs a computational process for designing high-affinity candidate antibodies that can bind to the epitope, and screens and verifies the computationally designed candidate antibodies through verification experiments, ultimately obtaining a target antibody that can bind to the target antigen epitope. Therefore, the target antibody computationally designed in this application targets an epitope related to the biological function specified by the user, the computational design process is precise and controllable, and the target antibody designed can be used in different fields such as the treatment of different diseases, as reagents and consumables in life sciences, and the development of diagnostic reagents.

[0065] At the same time, this application is suitable for the discovery of target antibodies for difficult-to-express, low immunogenic, and highly homologous antigens and epitopes, and unlike the antibody design method based on physical energy functions, this application generates fully human antibodies, which can avoid problems such as immunogenicity and developability, and theoretically ensures better drugability. And the AI ​​antibody design process of this application is used to design antibodies from scratch and combine them with experimental verification and screening. After the computing platform matures, the cycle is much shorter than traditional antibody discovery (1 to 2 months VS 4 to 6 months). Moreover, this application does not require animal immunization, which greatly reduces animal deaths and is beneficial to animal protection.

[0066] Referring to Figures 2 and 3 , in one example, the present application provides a method for de novo antibody design, comprising:

[0067] 010: Obtain the structure and epitope residue sequence of the target antigen.

[0068] In the present application, the target antigen can be a protein, and the structure of the target antigen can be a structure predicted by experimental means (such as cryo-electron microscopy, X-ray diffraction) or computational methods (such as AlphaFold, etc.).

[0069] In this application, there is no limitation on the storage format of the target antigen structure. It can be stored in pdb format or cif format. For example, the structure of the antigen is 1ZVM.pdb, and the antigen epitope residue sequence can be the epitope residue sequence on any chain or multiple chains.

[0070] 020: Based on the structure and epitope residue sequence of the target antigen, a library of related antibodies that can potentially bind to the target antigen epitope is generated. The related antibody library contains the structure and sequence of the related antibodies.

[0071] The number of related antibodies in the related antibody library in this application is in the tens of millions, for example, 1000w level. Related antibodies are related to the target antigen epitope, so that a large number of related antibodies can be generated, so that the related antibodies have diversity, and mainly the diversity of the CDR region, so that there are more potential related antibodies that can bind to the target antigen epitope in the related antibody library. And after generating a large number of related antibody sequences, data support is provided for the next step of narrowing the scope to antibodies that bind to specific epitopes. The related antibody sequences output by this application can be antibody light and heavy chain sequences (sequence format is [VH; VL], such as:

[0072] Since the present application designs a co-light chain antibody, it is used in the development of polyclonal antibodies to improve the efficiency of polyclonal antibody assembly, reduce light and heavy chain mismatches, and reduce the difficulty of chemical composition production and CMC control.

[0073] 030: Based on the structure and epitope residue sequence of the target antigen, as well as the structure and sequence of related antibodies in the related antibody library, the affinity of the sequence of the related antibody and the epitope residue sequence of the target antigen, as well as the structure of the related antibody and the structure of the target antigen are predicted based on the preset model to obtain the binding probability of the related antibody to the target antigen epitope.

[0074] This application uses a preset model to predict the affinity of the relevant antibody sequence and the target antigen epitope residue sequence, as well as the target antigen structure and the structure of the relevant antibody, to obtain the binding probability of the relevant antibody to the target antigen epitope. The prediction model in this application can be an antigen epitope-antibody affinity prediction model.

[0075] 031: From related antibodies with a preset binding probability, select antibodies with a preset proportion as candidate antibodies.

[0076] This application screens for antibodies with high affinity as candidate antibodies to roughly identify antibodies that can potentially bind to the target antigen epitope. For example, this application can select the 20%-30% sequences with the highest binding probability as candidate antibodies with high affinity. This application can also select the 20%-30% sequences with the second highest binding probability as candidate antibodies with high affinity.

[0077] In this way, the present application can screen out candidate antibodies with a higher probability of binding to the target antigen epitope from the relevant antibody library, so as to reduce the difficulty of designing subsequent target antibodies.

[0078] The candidate antibodies output in this application can be in the form of antibody light and heavy chain sequences + binding probability, for example:

[0079] 040: Perform humanization screening on candidate antibodies to exclude antibodies composed of non-human parts that are immunogenic, and retain antibodies composed of human parts that are non-immunogenic.

[0080] The antibodies obtained in this application through humanization screening are basically non-immunogenic, providing feasibility for subsequent development and reducing the cost of antibody optimization.

[0081] 041: Antibodies composed of human parts and without immunogenicity are designated as humanized antibodies, and a humanized antibody library is formed.

[0082] The number of humanized antibodies in the humanized antibody library of this applicant is in the millions, which is an order of magnitude smaller than that of related antibody libraries, greatly reducing the cost of subsequent verification experiments.

[0083] 042: If the accuracy of the preset model is high, a first preprocessing method is used to process a first preset number of candidate antibodies to obtain experimental antibody molecules and sequences.

[0084] If the candidate antibodies predicted by the prediction model have a high probability of actually binding to the target antibody epitope through subsequent experimental verification, the prediction model has a high accuracy. The first pretreatment method in this application includes recombinant expression, and the experimental antibody molecules and sequences are obtained by recombinantly expressing the candidate antibodies.

[0085] 043: If the accuracy of the preset model is low, a second preprocessing method is used to process the humanized antibodies in the humanized antibody library, and a second preset number of experimental antibody molecules and sequences are screened from the processed humanized antibodies.

[0086] If the prediction model predicts that the candidate antibody has a low probability of actually binding to the target antibody epitope through subsequent experimental verification, it means that the prediction model has low accuracy. The second preprocessing method of the present application includes:

[0087] The humanized antibody is divided into multiple fragments, which are connected by molecular biological methods, and the complete antibody sequence is constructed into a vector to form a library for display screening; the complete antibody sequence is screened in the library to obtain the processed humanized antibody.

[0088] For example, the humanized antibody framework region is gene-synthesized, a partial oligo pool of CDRH3 is synthesized, and a full-length antibody library is constructed using molecular biology methods. The antibody library is then screened using display technologies such as phage display and yeast display to obtain processed humanized antibodies. A second predetermined number of experimental antibody molecules and sequences are then screened from the processed humanized antibodies.

[0089] Because the candidate antibodies contain a large number of negative antibodies, the presence of negative antibodies will make subsequent validation experiments too costly and therefore need to be removed. After processing through the second pretreatment method, antibodies that may bind to the target antigen are found in the humanized library, narrowing the scope of subsequent experimental verification.

[0090] 050: Verify whether the experimental antibody molecule can bind to the target antigen epitope residue, and screen out the experimental antibody molecule that can bind to the target antigen epitope residue as the target antibody.

[0091] The present application can verify whether the experimental antibody molecule can block the binding of known antibodies / ligands to target antigen proteins by means of wet experiments such as ELISA / SPR / BLI, to verify whether the binding epitope of the experimental antibody molecule to the target antigen meets the design prediction, and the experimental antibody molecule that can block the binding of known antibodies / ligands to the target antigen protein is used as the target antibody. The complex structure of the experimental antibody molecule and the target antigen can also be obtained by the method of cryo-electron microscopy to verify whether the binding epitope of the experimental antibody molecule to the target antigen meets the design prediction, and the experimental antibody molecule that meets the design prediction is used as the target antibody. And the cell function experiment related to the antigen function can also be used to detect whether the experimental antibody molecule has the expected biological function, and the disease animal model related to the antigen function can also be used to detect whether the experimental antibody molecule can treat the corresponding disease in the animal model.

[0092] The present application can also verify whether the epitope bound to the target antigen by the experimental antibody molecule is correct through methods such as CryoEM and HDX, thereby finding the target functional antibody that can bind to the target epitope, which is more helpful in finding and confirming the antibody containing the target function, and can obtain dozens of target antibody sequences that bind to the correct epitope.

[0093] 060: If the affinity of the candidate antibody does not reach the preset threshold, the candidate antibody is subjected to affinity maturation to obtain a mature antibody with an affinity reaching the preset threshold.

[0094] Perform affinity maturation on candidate antibodies to simulate the evolutionary process of affinity maturation of antibodies in mice, enabling faster and more accurate affinity maturation of known candidate antibodies.

[0095] The present application characterizes affinity by binding probability, and the preset threshold may be the binding probability. The preset threshold is set according to the actual target antigen, for example, it may be 0.8 or 0.9.

[0096] Optionally, after obtaining the mature antibody, the present application may further include:

[0097] The candidate antibodies are screened for humanization to exclude antibodies composed of non-human parts that are immunogenic, and retain antibodies composed of human parts that are non-immunogenic.

[0098] Antibodies composed of human parts and without immunogenicity are referred to as humanized antibodies, and a humanized antibody library is formed.

[0099] If the accuracy of the preset model is high, a first pre-processing method is used to process a first preset number of candidate antibodies to obtain experimental antibody molecules and sequences. If the accuracy of the preset model is low, a second pre-processing method is used to process the humanized antibodies in the humanized antibody library, and a second preset number of experimental antibody molecules and sequences are screened from the processed humanized antibodies. It is verified whether the experimental antibody molecules can bind to the target antigen epitope residues, and the experimental antibody molecules that can bind to the target antigen epitope residues are screened as target antibodies.

[0100] Through the above method, target antibodies with higher affinity are screened.

[0101] The present application also provides a device for de novo antibody design.

[0102] Referring to FIG4 , the present application provides a device for designing antibodies from scratch, comprising:

[0103] The acquisition module 40 is used to obtain the structure and epitope residue sequence of the target antigen.

[0104] The antibody generation module 41 is used to generate a library of relevant antibodies that can potentially bind to the target antigen epitope based on the structure and epitope residue sequence of the target antigen. The relevant antibody library contains the structure and sequence of relevant antibodies. The antibody generation module 41 of the present application is configured with a sequence- and structure-based antibody pre-training and antibody generation model.

[0105] The affinity prediction module 42 is used to predict the affinity of the relevant antibodies for the target antigen epitope based on the structure and epitope residue sequence of the target antigen, as well as the structure and sequence of the relevant antibodies in the relevant antibody library, and screen out candidate antibodies that meet the preset affinity conditions. The affinity prediction module 42 of the present application is configured with an antigen epitope-antibody affinity prediction model.

[0106] The preprocessing module 43 is used to select a corresponding preprocessing method to process the candidate antibodies according to the accuracy of the preset model to obtain experimental antibody molecules and sequences. The experimental antibody molecules can be used as experimental subjects. The experiment is used to verify whether the experimental antibody molecules can bind to the target antigen epitope residues and screen out the experimental antibody molecules that can bind to the target antigen epitope residues as target antibodies.

[0107] In the present application, the affinity prediction module 42 includes a prediction module and a screening module. The prediction module is used to predict the affinity of the sequence of the relevant antibody with the epitope residue sequence of the target antigen, and the structure of the relevant antibody with the structure of the target antigen based on a preset model according to the structure and epitope residue sequence of the target antigen, as well as the structure of the relevant antibody with the structure of the target antigen; the screening module is used to screen out antibodies with a preset proportion as alternative antibodies from the relevant antibodies with a preset binding probability.

[0108] This application targets a user-specified target antigen epitope, designs a computational process from scratch for candidate antibodies with high affinity that can bind to the epitope, and screens and verifies the computationally designed candidate antibodies through verification experiments, ultimately obtaining a target antibody that can bind to the target antigen epitope. Therefore, the target antibody computationally designed in this application targets the epitope related to the biological function specified by the user. The computational design process is precise and controllable, and the target antibody designed can be used in different fields such as the treatment of different diseases, as reagents and consumables in life sciences, and the development of diagnostic reagents.

[0109] The antibody de novo design apparatus of the present application further includes a humanization screening module, which is used to perform humanization screening on candidate antibodies to obtain humanized antibodies and form a humanized antibody library. The humanization screening module of the present application is configured with an antibody humanization model.

[0110] The antibody de novo design apparatus of the present application also includes an affinity maturation module, which is used to perform affinity maturation on candidate antibodies to obtain mature antibodies with an affinity that reaches a preset threshold. The affinity maturation module of the present application is configured with an evolution-based antibody affinity maturation model.

[0111] Regarding the above-mentioned device, the specific manner in which each module performs operations has been described in detail in the related method and will not be elaborated again here.

[0112] The present application also provides an electronic device and a storage medium.

[0113] Referring to Figure 5, the electronic device 500 includes a memory 510 and a processor 520. The antibody de novo design method of the present application can be implemented by the electronic device 500 of the present application. Among them, the memory 510 stores executable code, and the processor 520 is used to obtain the structure and epitope residue sequence of the target antigen; based on the structure and epitope residue sequence of the target antigen, a related antibody library that can potentially bind to the target antigen epitope is generated, and the related antibody library contains the structure and sequence of the related antibodies; based on the structure and epitope residue sequence of the target antigen, as well as the structure and sequence of the related antibodies in the related antibody library, the affinity of the related antibody to the target antigen epitope is predicted based on the preset model, and the candidate antibodies that meet the preset affinity conditions are screened out; based on the accuracy of the preset model, the corresponding preprocessing method is selected to process the candidate antibodies to obtain the experimental antibody molecules and sequences.

[0114] The processor 520 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0115] The memory 510 may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage. ROM may store static data or instructions required by the processor 520 or other modules of the computer. The permanent storage may be a readable and writable storage device. The permanent storage may be a non-volatile storage device that retains stored instructions and data even when the computer is powered off. In some embodiments, the permanent storage device uses a large-capacity storage device (e.g., a magnetic or optical disk, flash memory) as the permanent storage device. In other embodiments, the permanent storage device may be a removable storage device (e.g., a floppy disk, optical drive). The system memory may be a readable and writable storage device or a volatile readable and writable storage device, such as dynamic random access memory. The system memory may store some or all instructions and data required by the processor during operation. In addition, the memory 510 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (e.g., DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and magnetic disks and / or optical disks may also be used. In some embodiments, the memory 510 may include a readable and / or writable removable storage device, such as a compact disc (CD), a read-only digital versatile disc (e.g., DVD-ROM, dual-layer DVD-ROM), a read-only Blu-ray disc, an ultra-density optical disc, a flash memory card (e.g., SD card, mini SD card, Micro-SD card, etc.), a magnetic floppy disk, etc. Computer-readable storage media do not include carrier waves and transient electronic signals transmitted wirelessly or wired.

[0116] The memory 510 stores executable codes. When the executable codes are processed by the processor 1020 , the processor 520 may execute part or all of the above-mentioned methods.

[0117] In addition, the method according to the present application may also be implemented as a computer program or a computer program product, which includes computer program code instructions for executing some or all of the steps in the above method of the present application.

[0118] Alternatively, the present application can also be implemented as a computer-readable storage medium (or non-transitory machine-readable storage medium or machine-readable storage medium) on which executable code (or computer program or computer instruction code) is stored. When the executable code (or computer program or computer instruction code) is executed by a processor of an electronic device (or server, etc.), the processor executes part or all of the steps of the above-mentioned method according to the present application.

[0119] In the description of this specification, the descriptions with reference to the terms "above", "wherein", "understandably", "further" and the like are intended to mean that the specific features, structures, materials or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are mutually inconsistent.

[0120] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code that includes one or more executable requests for implementing a specific logical function or step of a process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed in a different order than shown or discussed, including performing functions in a substantially simultaneous manner or in a reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0121] The embodiments of the present application have been described above. The above description is illustrative and not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is selected to best explain the principles of the embodiments, their practical applications, or improvements to the technology in the market, or to enable other persons skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for de novo antibody design, characterized in that: include: Obtain the structure and epitope residue sequence of the target antigen; According to the structure and epitope residue sequence of the target antigen, generating a library of related antibodies that can potentially bind to the epitope of the target antigen, wherein the library of related antibodies comprises the structure and sequence of related antibodies; According to the structure and epitope residue sequence of the target antigen, and the structure and sequence of the related antibodies in the related antibody library, the affinity of the related antibodies to the target antigen epitope is predicted based on a preset model, and candidate antibodies that meet the preset affinity conditions are screened out; According to the accuracy of the preset model, a corresponding preprocessing method is selected to process the candidate antibody to obtain an experimental antibody molecule and sequence; Verify whether the experimental antibody molecule can bind to the target antigen epitope residue, and screen out the experimental antibody molecule that can bind to the target antigen epitope residue as the target antibody.

2. The method according to claim 1, characterized in that The method predicts the affinity of the relevant antibodies to the target antigen epitope based on a preset model according to the structure and epitope residue sequence of the target antigen, and the structure and sequence of the relevant antibodies in the relevant antibody library, and screens out candidate antibodies that meet the preset affinity conditions, including: According to the structure and epitope residue sequence of the target antigen, and the structure and sequence of the related antibodies in the related antibody library, affinity prediction is performed on the sequence of the related antibody and the epitope residue sequence of the target antigen, and the structure of the related antibody and the structure of the target antigen based on a preset model to obtain the binding probability of the related antibody and the target antigen epitope; From the related antibodies whose binding probability is a preset probability, antibodies with a preset proportion are screened out as candidate antibodies.

3. The method according to claim 2, characterized in that The step of screening out antibodies with a preset proportion from the related antibodies with a preset binding probability as candidate antibodies comprises: From the sequences with the highest or second highest binding probability, 20%-30% of the sequences are selected as candidate antibodies.

4. The method according to claim 1, characterized in that Also includes: The candidate antibodies are subjected to humanization screening to obtain humanized antibodies, and a humanized antibody library is formed.

5. The method according to claim 4, characterized in that The step of performing humanization screening on the candidate antibodies to obtain humanized antibodies and forming a humanized antibody library comprises: Performing humanization screening on the candidate antibodies, eliminating antibodies composed of non-human parts that are immunogenic, and retaining antibodies composed of human parts that are non-immunogenic; The non-immunogenic antibodies composed of human parts are regarded as humanized antibodies, and a humanized antibody library is formed.

6. The method according to claim 2, characterized in that According to the accuracy of the preset model, a corresponding preprocessing method is selected to process the candidate antibody to obtain the experimental antibody molecule and sequence, including: If the accuracy of the preset model is high, a first preprocessing method is used to process a first preset number of candidate antibodies to obtain experimental antibody molecules and sequences; If the accuracy of the preset model is low, a second preprocessing method is used to process the humanized antibodies in the humanized antibody library, and a second preset number of experimental antibody molecules and sequences are screened from the processed humanized antibodies.

7. The method according to claim 6, characterized in that The first pretreatment method includes recombinant expression.

8. The method according to claim 6, characterized in that The second pretreatment method comprises: Dividing the humanized antibody into multiple fragments, connecting the fragments by molecular biological means, and constructing the complete antibody sequence into a vector to form a library for display screening; The complete antibody sequence is screened in the library to obtain the processed humanized antibody.

9. The method according to claim 1, characterized in that: Also includes: If the affinity of the candidate antibody does not reach the preset threshold, the candidate antibody is subjected to affinity maturation treatment to obtain a mature antibody whose affinity reaches the preset threshold.

10. The method according to claim 1, characterized in that Also includes: Verify whether the epitope to which the experimental antibody molecule binds to the target antigen is the target epitope, and screen out the target functional antibody that can bind to the target epitope of the target antigen.

11. A device for de novo antibody design, characterized in that: include: An acquisition module, used to obtain the structure and epitope residue sequence of the target antigen; An antibody generation module, used to generate a library of related antibodies that can potentially bind to the target antigen epitope according to the structure and epitope residue sequence of the target antigen, wherein the library of related antibodies contains the structure and sequence of related antibodies; An affinity prediction module, used to predict the affinity of the relevant antibody to the target antigen epitope based on a preset model according to the structure and epitope residue sequence of the target antigen, as well as the structure and sequence of the relevant antibodies in the relevant antibody library, and screen out candidate antibodies that meet the preset affinity conditions; The preprocessing module is used to select a corresponding preprocessing method to process the candidate antibody according to the accuracy of the preset model to obtain an experimental antibody molecule and sequence. The experimental antibody molecule can be used as an experimental object. The experiment is used to verify whether the experimental antibody molecule can bind to the Target antigen epitope residues, and screen out experimental antibody molecules that can bind to the target antigen epitope residues as target antibodies.

12. The device according to claim 11, characterized in that The affinity prediction module comprises: A prediction module, for predicting affinity between the sequence of the relevant antibody and the epitope residue sequence of the target antigen, and between the structure of the relevant antibody and the structure of the target antigen based on a preset model, according to the structure and epitope residue sequence of the target antigen, and the structure and sequence of the relevant antibody in the relevant antibody library, to obtain the binding probability of the relevant antibody and the epitope of the target antigen; The screening module is used to screen out antibodies with a preset proportion as candidate antibodies from the relevant antibodies with a preset binding probability.

13. The device according to claim 11, characterized in that The device also includes a humanization screening module, which is used to perform humanization screening on the candidate antibodies to obtain humanized antibodies and form a humanized antibody library.

14. An electronic device, characterized in that: include: processor; as well as A memory having executable codes stored thereon, which, when executed by the processor, causes the processor to execute the method according to any one of claims 1 to 10.

15. A computer-readable storage medium having executable codes stored thereon, which, when executed by a processor of an electronic device, causes the processor to execute the method according to any one of claims 1 to 10.

Images