Method, device and equipment for generating antibody based on human antibody skeleton library and storage medium
By screening and optimizing antibodies in the human antibody backbone library, the problems of inefficient and high cost of antibody design in traditional technology are solved, and the efficient binding of antibodies to specific epitopes of antigens is achieved, and the accuracy and efficiency of antibody drug development is improved. .
Patent Information
- Application Number
- CN202410716252.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-04
- Publication Date
- 2024-11-08
AI Technical Summary
Traditional hybridoma technology is inefficient and costly to obtain all-human antibodies. How to design antibodies that bind specific epitopes of antigens based on the antibody skeleton has become an urgent problem.
By obtaining the structure and sequence of the target antigen and the antibody structure and sequence in the human antibody backbone library, candidate antibodies that meet the preset conditions are selected and docking posture adjustment is performed. Sequence optimization and structure of the CDR region is combined with the preset algorithm. Optimization, use a combination algorithm to perform multi-angle calculations, and filter out the optimized docking structure and sequence that meet the conditions.
The efficiency and accuracy of antibody design are improved, and the generated target antibodies can stably bind to specific epitopes of antigens, reduce the error rate of subsequent experiments and improve the success rate of antibody drug development.
Smart Images

Figure CN118918981A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of bioinformatics technology, and in particular to a method, device, equipment and storage medium for generating antibodies based on a human antibody framework library. Background Art
[0002] Antibodies play an increasingly important role in the treatment of a variety of diseases due to their high targeting and specificity. Fully humanized antibodies are considered to be the main direction of future antibody drug development due to their advantages such as low immunogenicity and high affinity.
[0003] Traditional hybridoma technology is the main means of obtaining fully human antibodies, but it is inefficient and costly. In recent years, with the development of structural biology and computer-aided drug design technology, it has become possible to use computers to design antibody molecules from a theoretical level, which is expected to break through the bottleneck of traditional technology and improve the efficiency of human antibody design.
[0004] Among the related technologies, the technology of designing small proteins based on the skeleton is already quite mature, and the success rate of binding to specific epitopes is also very high.
[0005] In view of this, how to design antibodies that can bind to specific antigen epitopes based on the antibody skeleton has become an urgent problem that needs to be solved. 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 generating antibodies based on a human antibody framework library, which can optimize the CDR region of the antibody based on the human antibody framework library to generate human antibodies that can bind to specific antigen epitopes.
[0007] The first aspect of the present application provides a method for generating antibodies based on a human antibody framework library, comprising:
[0008] Obtain the structure and sequence of the target antigen, the target epitope region, and the structure and sequence of the human antibodies in the human antibody framework library;
[0009] According to the sequence of the target antigen, the target epitope region, and the sequence of the human antibodies in the human antibody framework library, human antibodies that meet the first preset condition are screened from the human antibody framework library as candidate antibodies;
[0010] Based on the target epitope region, docking the candidate antibody with the target antigen, adjusting the docking posture of the candidate antibody, and obtaining a docking structure;
[0011] Based on a preset algorithm, sequence optimization is performed on the CDR region of the antibody in each docking structure and structure optimization is performed on the binding interface to obtain multiple optimized docking structures and antibody sequences;
[0012] Based on a preset combination algorithm, a plurality of the optimized docking structures and antibody sequences are calculated from multiple angles to screen out the optimized docking structures and antibody sequences that meet the preset combination screening conditions, and the structure and sequence of the antibody in the optimized docking structure are used as the structure and sequence of the target antibody.
[0013] As an optional embodiment, the method of screening out human antibodies that meet the first preset condition from the human antibody framework library according to the sequence of the target antigen, the target epitope region, and the sequence of the human antibodies in the human antibody framework library as candidate antibodies includes:
[0014] Predicting a first complex structure of the target antigen and the human antibody based on the sequence of the target antigen and the sequence of the human antibody;
[0015] According to the binding interface of the first complex structure and the target epitope region, a first complex structure that meets the first preset condition is screened, and the human antibodies in the first complex structure corresponding to the human antibody framework library are taken as candidate antibodies.
[0016] As an optional embodiment, docking the candidate antibody with the target antigen based on the target epitope region, adjusting the docking posture of the candidate antibody to obtain a docking structure includes:
[0017] According to the structure of the target antigen and the target epitope region, low-precision docking antigen residues and high-precision docking antigen residues are screened from the residues of the target antigen structure;
[0018] Based on the low-precision docked antigen residues, the candidate antibody is docked with the target antigen with low precision, and candidate antibodies that satisfy shape complementarity and their relative position data with the target antigen are screened and obtained;
[0019] Based on the antigen residues for high-precision docking, the candidate antibody satisfying shape complementarity and its relative position data with the target antigen, the candidate antibody satisfying shape complementarity is docked with the target antigen with high precision. During the high-precision docking process, the docking posture of the candidate antibody satisfying shape complementarity is adjusted to obtain a docking structure.
[0020] As an optional embodiment, the step of screening low-precision docking antigen residues and high-precision docking antigen residues from the residues of the target antigen structure according to the structure of the target antigen and the target epitope region comprises:
[0021] Analyzing the hydrophobicity of all residues in the structure of the target antigen, and screening out residues with a hydrophobicity greater than a third threshold as candidate residues;
[0022] Selecting clustered residues from the residues contained in the target epitope region, and selecting a first preset number of residues from the clustered residues with the candidate residues as the preferred candidate residues as antigen residues for low-precision docking;
[0023] The low-precision docked antigen residues and residues in their neighborhood are combined into a second preset number of high-precision docked antigen residues, wherein the residues in the neighborhood of the low-precision docked antigen residues belong to the residues contained in the target epitope region.
[0024] As an optional embodiment, the low-precision docking of the candidate antibody with the target antigen based on the low-precision docked antigen residues, and screening to obtain candidate antibodies that satisfy shape complementarity and their relative position data with the target antigen, includes:
[0025] mutating all the residues on the main chain of the candidate antibody to valine;
[0026] The posture of the candidate antibody is adjusted to screen out the candidate antibody whose main chain can be shape-complementary with the target epitope region of the target antigen, and obtain the candidate antibody that satisfies shape complementarity and its relative position data with the target antigen.
[0027] As an optional embodiment, based on the antigen residues for high-precision docking and the candidate antibody satisfying shape complementarity and the relative position data between the candidate antibody satisfying shape complementarity and the target antigen, the candidate antibody satisfying shape complementarity is docked with the target antigen with high precision, and during the high-precision docking process, the docking posture of the candidate antibody satisfying shape complementarity is adjusted to screen a docking structure satisfying the second preset condition, including:
[0028] Based on the antigen residue docked with high precision, generating a preset side chain that interacts with the antigen residue near the designated antigen residue docked with high precision;
[0029] Calculating the binding energy data of the preset side chain and the target antigen, and storing the preset side chain and the corresponding binding energy data to obtain a side chain matching library;
[0030] Based on the candidate antibody satisfying shape complementarity and the relative position data between the candidate antibody satisfying shape complementarity and the target antigen, the docking posture of the candidate antibody satisfying shape complementarity is adjusted, and the side chain of the candidate antibody satisfying shape complementarity is matched with the preset side chain in the side chain matching library to obtain the binding energy data corresponding to the side chain in the candidate antibody satisfying shape complementarity that matches the preset side chain in the side chain matching library;
[0031] According to the binding energy data of the side chains in the candidate antibodies satisfying shape complementarity, the side chains in the candidate antibodies satisfying shape complementarity are replaced with preset side chains or retained to obtain a docking structure.
[0032] As an optional embodiment, after obtaining the docking structure, the method further includes:
[0033] The FR regions of the candidate antibody and the antibody in the docking structure are compared, and the residues in the FR region of the antibody in the docking structure are mutated to the same residues as those in the FR region of the candidate antibody.
[0034] As an optional embodiment, after obtaining the docking structure, the method further includes the step of screening the docking structure that meets the second preset condition:
[0035] analyzing the interacting residues on the antibody in the docking structure and the target antigen;
[0036] Calculating the proportion of the interacting residues located in the CDR region of the antibody in the docking structure, and screening out docking structures having a proportion greater than a fourth threshold;
[0037] Calculating the optimization screening index of the docking structure after screening, wherein the optimization screening index includes at least one of: binding free energy of the antigen-antibody interface, hydrophobic change before and after antigen-antibody binding, and hydrophobic contact area between antigen and antibody;
[0038] The docking structure whose optimized screening index meets the preset screening condition is screened out.
[0039] As an optional embodiment, based on a preset algorithm, the CDR region of the antibody in each docking structure is sequence optimized and the binding interface is structurally optimized to obtain multiple optimized docking structures and antibody sequences, including:
[0040] Pre-mutating the residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure according to the frequencies of occurrence of different types of amino acids at different positions in the CDRH1 and CDRH2 regions of the natural antibody;
[0041] Determining whether to mutate or not mutate residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure according to energy changes before and after the pre-mutation of the docking structure;
[0042] Premutating the residues in the CDRH3 region of the antibody in the docking structure to preset residues other than cysteine and proline, and determining whether to mutate or not mutate the residues in the CDRH3 region of the antibody in the docking structure according to the energy change before and after the premutation of the docking structure;
[0043] According to the above method, the residues in the CDRH1, CDRH2 and CDRH3 regions of the antibody in the docking structure are iteratively mutated multiple times, and the binding interface of the docking structure after each mutation is structurally optimized to obtain an optimized docking structure;
[0044] The above process was repeated multiple times to obtain multiple optimized docking structures and antibody sequences.
[0045] As an optional embodiment, the method of performing multi-angle calculations on the plurality of optimized docking structures and antibody sequences based on a preset combination algorithm, screening out the optimized docking structures and antibody sequences that meet the preset combination screening conditions, and using the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody includes:
[0046] Calculating and obtaining a plurality of first screening indicators of the optimized docking structures, wherein the first screening indicator comprises at least one of: binding free energy of the antigen-antibody interface, a hole score of the antigen-antibody interface, a change in hydrophobicity before and after antigen-antibody binding, and a hydrophobic contact area between the antigen and antibody;
[0047] The optimized docking structure whose first screening index meets the first screening condition is screened out, and the structure and sequence of the antibody in the optimized docking structure are used as the structure and sequence of the target antibody.
[0048] As an optional embodiment, the method of performing multi-angle calculations on the plurality of optimized docking structures and antibody sequences based on a preset combination algorithm, screening out the optimized docking structures and antibody sequences that meet the preset combination screening conditions, and using the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody includes:
[0049] According to the structure of the target antigen and the sequence of the antibody in the optimized docking structure, the affinity of the target antigen and the antibody in the optimized docking structure is predicted, the optimized docking structure whose affinity meets the preset second screening condition is screened, and the structure and sequence of the antibody in the optimized docking structure are used as the structure and sequence of the target antibody.
[0050] As an optional embodiment, the method of performing multi-angle calculations on the plurality of optimized docking structures and antibody sequences based on a preset combination algorithm, screening out the optimized docking structures and antibody sequences that meet the preset combination screening conditions, and using the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody includes:
[0051] According to the sequence of the target antigen and the sequence of the antibody in the optimized docking structure, predicting a second complex structure of the target antigen bound to the antibody in the optimized docking structure, and a non-binding structure of the antibody in the optimized docking structure;
[0052] According to the second complex structure and the non-binding structure of the antibody in the optimized docking structure, a plurality of second screening indicators of the optimized docking structure are calculated, wherein the second screening indicator includes at least one of the following indicators:
[0053] Confidence of the binding interface of the second complex structure; Confidence of the relative positions of the antigen and the antibody in the second complex structure; RMSD value of the antibody structure in the optimized docking structure and the antibody structure in the second complex structure after the structure of the antigen in the optimized docking structure is superimposed with the structure of the antigen in the second complex structure; RMSD value of the CDRH3 region of the antibody in the optimized docking structure and the CDRH3 region of the antibody in the second complex structure; RMSD value of the CDRH3 region of the antibody structure in the second complex structure and the non-binding structure of the antibody in the optimized docking structure; overlap between the binding epitope of the antibody and the antigen in the second complex structure and the target epitope;
[0054] The optimized docking structure whose second screening index satisfies the third screening condition is screened out, and the structure and sequence of the antibody in the optimized docking structure are used as the structure and sequence of the target antibody.
[0055] The second aspect of the present application provides a device for generating antibodies based on a human antibody framework library, comprising:
[0056] An acquisition module, used to obtain the structure and sequence of the target antigen, the target epitope region, and the structure and sequence of the human antibody in the human antibody framework library;
[0057] A prediction module, for screening human antibodies that meet a first preset condition from the human antibody framework library as candidate antibodies based on the sequence of the target antigen, the target epitope region, and the sequence of the human antibodies in the human antibody framework library;
[0058] A docking module, used for docking the candidate antibody with the target antigen based on the target epitope region, adjusting the docking posture of the candidate antibody, and screening to obtain a docking structure that meets a second preset condition;
[0059] An optimization module, for performing sequence optimization on the CDR region of the antibody in each docking structure and structural optimization on the binding interface based on a preset algorithm, to obtain a plurality of optimized docking structures and antibody sequences;
[0060] A screening module is used to perform multi-angle calculations on the plurality of optimized docking structures based on a preset combination algorithm, screen out the optimized docking structures and sequences that meet the preset combination screening conditions, and use the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody.
[0061] A third aspect of the present application provides an electronic device, including:
[0062] Processor; and
[0063] The memory stores executable codes thereon, and when the executable codes are executed by the processor, the processor is caused to execute the method as described above.
[0064] A fourth aspect of the present application provides a computer-readable storage medium having executable code stored thereon. When the executable code is executed by a processor of an electronic device, the processor is caused to execute the method as described above.
[0065] The technical solution provided by this application may have the following beneficial effects:
[0066] The present application is based on a human antibody framework library, and antibodies that meet the first preset condition are preliminarily screened out from the human antibody framework library as candidate antibodies; then the candidate antibody is docked with the target antigen, the docking posture of the candidate antibody is adjusted, and a docking structure with a more stable docking posture and meeting the second preset condition is screened; then the CDR region of the antibody in the docking structure is sequence optimized, and the antigen-antibody binding interface in the docking structure is structurally optimized to meet the diversity of the antibody CDR region and ensure the stability of the antibody CDR region structure; finally, through a preset combination algorithm, a reasonable screening method and strategy are set to perform multi-angle calculations on multiple optimized docking structures, so as to achieve a comprehensive evaluation and optimization of multiple optimized docking structures, and the screened optimized docking structure meets the evaluation of multiple conditions, so that the obtained target antibody is both a human antibody and has a high probability of binding to the target epitope region, greatly improving the accuracy of the subsequent experimental verification conclusions.
[0067] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0068] The above and other objects, features and advantages of the present application will become more apparent by describing in more detail 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.
[0069] Figure 1It is a schematic diagram of a process for generating antibodies based on a human antibody framework library as shown in an embodiment of the present application;
[0070] Figure 2 is a schematic flow chart of a method for generating antibodies based on a human antibody framework library according to an embodiment of the present application (including S10 to S37);
[0071] Figure 3 is a schematic flow chart of a method for generating antibodies based on a human antibody framework library according to an embodiment of the present application (including S40 to S55);
[0072] Figure 4 It is a schematic diagram of the structure of a device for generating antibodies based on a human antibody framework library shown in an embodiment of the present application;
[0073] Figure 5 It is a schematic diagram of the structure of an electronic device shown in an embodiment of the present application. DETAILED DESCRIPTION
[0074] The embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, 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.
[0075] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.
[0076] It should be understood that although the terms "first", "second", "third", etc. may be used in this application to describe various information, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of this application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.
[0077] Among the related technologies, the technology of designing small proteins based on the skeleton is already quite mature, and the success rate of binding to specific epitopes is also very high.
[0078] In view of this, how to design antibodies that can bind to specific antigen epitopes based on the antibody skeleton has become an urgent problem that needs to be solved.
[0079] In response to the above problems, an embodiment of the present application provides a method for generating antibodies based on a human antibody framework library, which can optimize the CDR region of the antibody based on the human antibody framework library to generate human antibodies that can bind to specific epitopes of antigens.
[0080] The technical solution of the embodiments of the present application is described in detail below with reference to the accompanying drawings.
[0081] Figure 1 It is a schematic diagram of the process of generating antibodies based on a human antibody framework library shown in an embodiment of the present application.
[0082] See also Figure 1 The present invention provides a method for generating antibodies based on a human antibody framework library, comprising steps S1 to S5:
[0083] S1: Obtain the structure and sequence of the target antigen, the target epitope region, and the structure and sequence of the human antibodies in the human antibody backbone library.
[0084] The antigen in the present application embodiment refers to all substances that can stimulate the body to produce specific immune responses (humoral immunity and cellular immunity). The part where the antigen binds to the antibody is called the antigen epitope, and the target epitope region can refer to the epitope region of the target antigen that contains the target surface residues.
[0085] The structure of the target antigen in the embodiment of the present application can be obtained by the following method: the structure of the target antigen and its receptor complex is obtained from the PDB database, and the structure of the complex is repaired, and the structure of the target antigen is separated therefrom. It can also be a structure predicted by experimental means (such as cryo-electron microscopy, X-ray diffraction) or computational methods (such as AlphaFold, etc.). It should also be noted that the structure of the target antigen can also be modeled using the modeling method of the relevant technology, and this application does not limit this.
[0086] The sequence of the target antigen in the examples of the present application can be obtained from, for example, the TDC database, and the present application is not limited to this.
[0087] The human antibody framework library in the embodiment of the present application can be: using different antigens to immunize mice with co-light chains, isolating single B cells from the immunized mice, obtaining a large number of antibody heavy chain sequences through PCR and high-throughput sequencing technology, combining a large number of heavy chain sequences with co-light chains to form a large number of human antibody sequences, and obtaining the structure of the corresponding human antibody based on sequence prediction, and constructing a human antibody framework library using the structure and sequence of these human antibodies. Among them, the human antibodies in the human antibody framework library retain the affinity and specificity of mouse antibodies and reduce their heterology, which is beneficial for application in the human body.
[0088] S2: Based on the sequence of the target antigen, the target epitope region, and the sequence of the human antibodies in the human antibody framework library, human antibodies that meet the first preset condition are screened from the human antibody framework library and used as candidate antibodies.
[0089] In some embodiments, the embodiments of the present application can screen candidate antibodies from the human antibody framework library in the following manner: predict the first complex structure of the target antigen and the human antibody based on the sequence of the target antigen and the sequence of the human antibody; screen the first complex structure that meets the first preset condition based on the binding interface and the target epitope region of the first complex structure, and use the human antibody in the first complex structure that corresponds to the human antibody framework library as the candidate antibody. The antibody structure in the corresponding antibody framework library can be used in the subsequent docking process.
[0090] S3: Based on the target epitope region, the candidate antibody is docked with the target antigen, the docking posture of the candidate antibody is adjusted, and a docking structure that meets the second preset condition is screened.
[0091] In some embodiments, the embodiments of the present application can obtain a docking structure in the following manner: based on the structure of the target antigen and the target epitope region, the target antigen residues for low-precision docking and the target antigen residues for high-precision docking are screened from the residues of the target antigen structure; based on the target antigen residues for low-precision docking, the candidate antibody is docked with the target antigen with low precision, and the candidate antibody that satisfies shape complementarity and the relative position data with the target antigen are screened; based on the target antigen residues for high-precision docking and the candidate antibody that satisfies shape complementarity and the relative position data with the target antigen, the candidate antibody that satisfies shape complementarity is docked with the target antigen with high precision, and the docking posture of the candidate antibody that satisfies shape complementarity is adjusted during the high-precision docking process to obtain a docking structure that meets the second preset condition.
[0092] S4: Based on the preset algorithm, the CDR region of the antibody in each docking structure is sequence optimized and the binding interface is structurally optimized to obtain multiple optimized docking structures and antibody sequences.
[0093] Since the antibody in the embodiment of the present application is a co-light chain antibody, only the CDR region of the antibody heavy chain can be optimized in step S4. The multiple optimized docking structures obtained after optimization meet the characteristics of diversity and stability of the CDR region.
[0094] In some embodiments, the embodiments of the present application can optimize the CDR region in the following manner: pre-mutate the residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure according to the frequency of occurrence of different types of amino acids at different positions in the CDRH1 and CDRH2 regions of the natural antibody; determine whether to mutate or not the residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure according to the energy change before and after the pre-mutation of the docking structure; pre-mutate the residues in the CDRH3 region of the antibody in the docking structure to preset residues other than cysteine and proline, and determine whether to mutate or not the residues in the CDRH3 region of the antibody in the docking structure according to the energy change before and after the pre-mutation of the docking structure; iterate the residues in the CDRH1, CDRH2 and CDRH3 regions of the antibody in the docking structure multiple times according to the above method, and optimize the binding interface of the docking structure after each mutation to obtain an optimized docking structure; repeat the above process multiple times to obtain multiple optimized docking structures and antibody sequences.
[0095] S5: Based on the preset combination algorithm, multiple optimized docking structure antibody sequences are calculated from multiple angles to screen out the optimized docking structures and antibody sequences that meet the preset combination screening conditions, and the structure and sequence of the antibody in the optimized docking structure are used as the structure and sequence of the target antibody.
[0096] The embodiment of the present application presets a combination algorithm, sets a reasonable screening method and strategy to perform multi-angle calculations on multiple optimized docking structures, thereby achieving comprehensive evaluation and optimization of multiple optimized docking structures. The screened optimized docking structures meet the evaluation of multiple conditions, so that the obtained target antibodies are both human antibodies and have a high probability of binding to the target epitope region, greatly improving the accuracy of subsequent experimental verification conclusions.
[0097] In some implementations, the embodiments of the present application can screen out optimized docking structures that meet the preset combination screening conditions in the following manner:
[0098] Method 1: Calculate and obtain a first screening index of multiple optimized docking structures, wherein the first screening index includes at least one of: binding free energy of the antigen-antibody interface, pore score of the antigen-antibody interface, hydrophobicity change before and after antigen-antibody binding, and hydrophobic contact area between the antigen and antibody; screen out the optimized docking structure whose first screening index meets the first screening condition, and use the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody.
[0099] Method 2: According to the structure of the target antigen and the sequence of the antibody in the optimized docking structure, the affinity of the target antigen and the antibody in the optimized docking structure is predicted, and the optimized docking structure whose affinity meets the preset second screening condition is screened out, and the structure and sequence of the antibody in the optimized docking structure are used as the structure and sequence of the target antibody.
[0100] Method three: predicting the second complex structure of the target antigen and the antibody in the optimized docking structure, and the non-binding structure of the antibody in the optimized docking structure according to the sequence of the target antigen and the sequence of the antibody in the optimized docking structure; calculating the second screening index of the multiple optimized docking structures according to the second complex structure and the non-binding structure of the antibody in the optimized docking structure, wherein the second screening index includes at least one of the following indexes: the confidence of the binding interface of the second complex structure; the confidence of the relative position of the antigen and the antibody in the second complex structure; the RMSD value of the antibody structure in the optimized docking structure and the antibody structure in the second complex structure after the structure of the antigen in the optimized docking structure and the structure of the antigen in the second complex structure are superimposed; the RMSD value of the CDRH3 region of the antibody in the optimized docking structure and the CDRH3 region of the antibody in the second complex structure; the RMSD value of the CDRH3 region of the antibody structure in the second complex structure and the non-binding antibody structure in the optimized docking structure; the overlap between the antigen epitope bound to the antibody in the second complex structure and the target epitope; screening out the optimized docking structure whose second screening index meets the third screening condition, and taking the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody.
[0101] As can be seen from this example, the embodiment of the present application is based on a human antibody framework library, and antibodies that meet the first preset condition are screened from the human antibody framework library as candidate antibodies; the candidate antibody is then docked with the target antigen, the docking posture of the candidate antibody is adjusted, and a docking structure with a more stable docking posture and that meets the second preset condition is quickly screened; the CDR region of the antibody in the docking structure is then sequence optimized, and the antigen-antibody binding interface in the docking structure is structurally optimized to meet the diversity of the antibody CDR region and ensure the stability of the antibody CDR region structure; finally, through a preset combination algorithm, a reasonable screening method and strategy are set to perform multi-angle calculations on multiple optimized docking structures, so as to achieve a comprehensive evaluation and optimization of multiple optimized docking structures, and the screened optimized docking structure meets the evaluation of multiple conditions, so that the obtained target antibody is both a human antibody and has a high probability of binding to the target epitope region, greatly improving the accuracy of the subsequent experimental verification conclusions.
[0102] Figure 2is a schematic flow chart of a method for generating antibodies based on a human antibody framework library according to an embodiment of the present application (including S10 to S37); Figure 3 It is a schematic flow chart of a method for generating antibodies based on a human antibody framework library as shown in an embodiment of the present application (including S40 to S55).
[0103] See also Figure 2 and Figure 3 The present invention provides a method for generating antibodies based on a human antibody framework library, comprising the following steps:
[0104] S10: Obtain the structure and sequence of the target antigen, the target epitope region, and the sequence of the human antibody in the human antibody backbone library.
[0105] The antigen in the present application embodiment refers to all substances that can stimulate the body to produce specific immune responses (humoral immunity and cellular immunity). The part where the antigen binds to the antibody is called the antigen epitope, and the target epitope region can refer to the epitope region of the target antigen that contains the target surface residues.
[0106] The structure of the target antigen in the embodiment of the present application can be obtained by the following method: the structure of the target antigen and its receptor complex is obtained from the PDB database, and the structure of the complex is repaired, and the structure of the target antigen is separated therefrom. It can also be a structure predicted by experimental means (such as cryo-electron microscopy, X-ray diffraction) or computational methods (such as AlphaFold, etc.). It should also be noted that the structure of the target antigen can also be modeled using the modeling method of the relevant technology, and this application does not limit this.
[0107] The sequence of the target antigen in the examples of the present application can be obtained from, for example, the TDC database, and the present application is not limited to this.
[0108] The human antibody framework library in the embodiment of the present application can be: using different antigens to immunize mice with co-light chains, isolating single B cells from the immunized mice, obtaining a large number of antibody heavy chain sequences through PCR and high-throughput sequencing technology, combining a large number of heavy chain sequences with co-light chains to form a large number of human antibody sequences, and obtaining the structure of the corresponding human antibody based on sequence prediction, and constructing a human antibody framework library using the structure and sequence of these human antibodies. Among them, the human antibodies in the human antibody framework library retain the affinity and specificity of mouse antibodies and reduce their heterology, which is beneficial for application in the human body.
[0109] S20: Predicting the structure of the first complex of the target antigen and the human antibody based on the sequence of the target antigen and the sequence of the human antibody.
[0110] In the present embodiment, Aplphafold2 can be used to predict the structure of the complex of the target antigen and the human antibody. Aplphafold2 uses 5 sets of model weights by default, and each set of model weights predicts 5 structures, that is, 5x5 structures. In order to save the amount of calculation, the present embodiment can use 1 set of model weights to predict 1x1 structure.
[0111] S21: Screening out the first complex structure that meets the first preset condition according to the binding interface and the target epitope region of the first complex structure, and taking the human antibodies in the first complex structure that correspond to the human antibody framework library as candidate antibodies.
[0112] In some embodiments, the first preset condition in the embodiments of the present application may be that the confidence of the binding interface of the first complex structure is greater than a first threshold, and the overlap between the binding epitope region of the antigen and antibody in the first complex structure and the target epitope region is greater than a second threshold.
[0113] The confidence of the binding interface in the complex structure of the target antigen and the human antibody in the embodiment of the present application reflects the quality of the complex binding interface. The higher the confidence, the higher the credibility of the complex structure; the higher the overlap between the binding epitope of the antigen and the antibody and the target epitope region in the complex structure of the target antigen and the human antibody, the more likely the antibody in the complex structure is to bind to the target epitope region of the target antigen. Therefore, the credibility of the complex structure generated by the combination of the target antigen and the human antibody screened by the first preset condition is high, and the possibility of binding to the target epitope region is also higher.
[0114] Among them, the first threshold can be 80 or set according to business needs, and the second threshold can be 0.5 or set according to business needs, which is not limited in this application.
[0115] S30: According to the structure of the target antigen and the target epitope region, low-precision docking antigen residues and high-precision docking antigen residues are screened from the residues of the target antigen structure.
[0116] In some embodiments, the embodiments of the present application can screen out low-precision docked antigen residues and high-precision docked antigen residues in the following manner: analyze the hydrophobicity of all residues in the structure of the target antigen, and screen out residues with a hydrophobicity greater than a third threshold as candidate residues; select clustered residues from the residues contained in the target epitope region, and among the clustered residues, preferably screen a first preset number of residues as low-precision docked antigen residues with the candidate residues as the preferred residues; and combine the low-precision docked antigen residues and the residues in their neighborhood to form a second preset number of high-precision docked antigen residues, wherein the residues in the neighborhood of the low-precision docked antigen residues belong to the residues contained in the target epitope region.
[0117] Since the protein-protein interaction is mainly hydrophobic, the docking residues screened in the examples of the present application are all hydrophobic residues.
[0118] Among them, before analyzing the hydrophobicity of all residues in the structure of the target antigen, the structure of the target antigen can be optimized using a physical model to adjust the unreasonable structure of the target antigen.
[0119] Hydrophobicity can be determined by the spatial aggregation tendency SAP model for the SAP scores of all residues in the structure of the target antigen. The SAP score is usually between 0 and 5. The third threshold of the embodiment of the present application can select residues with a SAP score greater than 1 or other empirical values as candidate residues, and the present application does not impose any limitation on this.
[0120] Clustered residues refer to some residues that are relatively close to each other. Generally, the residues in the target epitope region are some residues that are relatively close to each other, but the range may be relatively large and the number of residues is relatively large. It is not necessary to dock all residues during docking, so clustered residues in the target epitope region can be selected for low-precision docking, and clustered residues can be selected and checked by visual structure. The first preset number can be set to 15 or other empirical values, which is not limited in this application.
[0121] When selecting antigen residues for high-precision docking, the number of antigen residues required is more than that for low-precision docking. Therefore, based on the selection of antigen residues for low-precision docking, further selection can be made within their neighborhood. The second preset number is greater than the first preset number and can be set to 30 or other empirical values. This application does not limit this.
[0122] S31: Based on the low-precision docked antigen residues, the candidate antibody is locally docked with the target antigen with low precision to screen for candidate antibodies that meet shape complementarity and their relative position data with the target antigen.
[0123] In some embodiments, the embodiments of the present application can perform low-precision docking in the following manner: mutate all the residues on the main chain of the candidate antibody to valine; adjust the posture of the candidate antibody, screen out the candidate antibodies whose main chain and the target epitope region of the target antigen are shape-complementary, and obtain the candidate antibodies that satisfy shape complementarity and their relative position data with the target antigen.
[0124] The low-precision docking in the embodiments of the present application mainly considers the shape complementarity between the candidate antibody and the target epitope region of the target antigen. Since the side chain of valine is very short, all the residues on the main chain of the candidate antibody can be mutated to valine, and there is no need to consider the side chain. The shape complementarity between the main chain of the candidate antibody and the target antigen can be mainly considered to achieve low-precision docking of the candidate antibody and the target antigen, and the candidate antibodies that meet the shape complementarity and their relative position data with the target antigen are screened.
[0125] S32: Based on the high-precision docked antigen residues and the candidate antibodies that satisfy shape complementarity and their relative position data with the target antigen, the candidate antibodies that satisfy shape complementarity are docked with the target antigen with high precision. During the high-precision docking process, the docking posture of the candidate antibodies that satisfy shape complementarity is adjusted to obtain a docked structure.
[0126] In some implementations, the embodiments of the present application can achieve high-precision docking by:
[0127] Based on the high-precision docked antigen residues, a preset side chain that interacts with the antigen residue is generated near the designated high-precision docked antigen residue; the binding energy data of the preset side chain and the target antigen are calculated, and the preset side chain and the corresponding binding energy data are stored to obtain a side chain matching library; based on the candidate antibodies that meet shape complementarity and their relative position data with the target antigen, the docking posture of the candidate antibodies that meet shape complementarity is adjusted, and the side chains of the candidate antibodies that meet shape complementarity are matched with the preset side chains in the side chain matching library to obtain the binding energy data corresponding to the side chains in the candidate antibodies that meet shape complementarity that match the preset side chains in the side chain matching library; according to the binding energy data of the side chains in the candidate antibodies that meet shape complementarity, the side chains in the candidate antibodies that meet shape complementarity are replaced with preset side chains or retained to obtain a docking structure, and the binding energy of the antigen and antibody can also be calculated for subsequent screening.
[0128] The relative position data between the candidate antibody and the target antigen that satisfy shape complementarity may include data of relative rotation and translation.
[0129] The purpose of the low-precision docking in the embodiment of the present application is to pre-fix the target epitope region of the candidate antibody and the target antigen to stabilize the candidate antibody and the target antigen; during high-precision docking, the candidate antibody screened out can directly bind to the target epitope region of the target antigen.
[0130] The embodiment of the present application can generate preset side chains that interact with the antigen residues around the antigen residues for high-precision docking of the target antigen in advance, and calculate the binding energy data of the preset side chains and the target antigen. The generated binding energy data of the preset side chains and the target antigen are equivalent to the binding energy data of the antibody side chains binding to the target antigen. The embodiment of the present application calculates and stores the binding energy of the side chains and the target antigen in advance. When performing high-precision docking, it is only necessary to match the side chains of the antibody with the stored side chains to quickly obtain the binding energy of the side chains of the antibody and the target antigen. There is no need to calculate the binding energy of each antibody side chain with the target antigen, which can greatly save the amount of calculation. Based on this, when the candidate antibody that meets shape complementarity is docked with the target antigen with high precision, the docking posture of the candidate antibody that meets shape complementarity can be adjusted on the basis of the candidate antibody that meets shape complementarity obtained by low-precision docking and its relative position data with the target antigen, and different docking postures are obtained. A side chain matching library that interacts with the antigen residue is generated in advance around the antigen residue that is docked with high precision, so that the side chain of the antibody can be compared with the side chain in the side chain matching library, and the binding energy data corresponding to the side chain that matches the preset side chain in the side chain matching library in the candidate antibody that meets shape complementarity can be obtained. According to the binding energy data of the side chain in the candidate antibody that meets shape complementarity, the antibody side chain with a binding energy data less than a threshold value (indicating that it is conducive to binding) can be retained, and the antibody side chain with a binding energy data greater than a threshold value (indicating that it is not conducive to binding) can be replaced with a preset side chain that matches in the side chain matching library to obtain a docking structure. Therefore, the antibody in the obtained docking structure can stably bind to the target epitope region of the target antigen. Finally, the binding energy of the antigen-antibody is calculated, and the antigen-antibody docking structure with the top 5000 energy ranking is screened for each candidate antibody.
[0131] In some embodiments, after obtaining the docking structure, the method further comprises:
[0132] S33: Compare the FR region of the candidate antibody and the antibody in the antigen-antibody complex structure obtained by docking, and mutate the residues in the FR region of the antibody in the docking structure to the same residues as those in the FR region of the candidate antibody.
[0133] Since in step S32, the side chains in the candidate antibodies that satisfy shape complementarity can be replaced with preset side chains according to the binding energy data of the side chains in the candidate antibodies that satisfy shape complementarity to obtain the docking structure, the replaced area cannot be controlled during the replacement, and the FR region of the antibody may be replaced, but the replacement of the FR region will bring some risks affecting expression, immunogenicity, etc. Therefore, the embodiment of the present application can mutate the residues in the FR region of the antibody in the docking structure that are different from those in the FR region of the candidate antibody to the same residues as the FR region of the candidate antibody (referred to as restoring mutations) to avoid the FR region mutation affecting the antibody expression and reducing the degree of humanization.
[0134] In some embodiments, after obtaining the docking structure, the step of screening the docking structure that meets the second preset condition is further included:
[0135] S34: Analyze the interacting residues on the antibody with the target antigen in the docked structure.
[0136] In the present embodiment, each residue on the antibody in the docking structure can be paired with each residue on the target antigen to form a residue pair, and the distance between the atoms included in each residue pair is calculated. The residue pairs of are taken as the interacting residue pairs, and the residues in the interacting residue pairs located on the antibody in the docking structure are taken as the interacting residues on the antibody in the docking structure with the target antigen.
[0137] S35: Calculate the proportion of the interacting residues located in the CDR region of the antibody in the docking structure, and screen out docking structures with a proportion greater than a fourth threshold.
[0138] Since the CDR region of the antibody generally binds to the target epitope region of the target antigen, the docking structure in which the interacting residues on the antibody and the target antigen are basically located in the CDR region is screened. The fourth threshold value can be set to 0.5 or other empirical values, which is not limited in this application.
[0139] S36: Calculate the optimization screening index of the docking structure after screening, wherein the optimization screening index includes at least one of: binding free energy of the antigen-antibody interface, hydrophobicity change before and after antigen-antibody binding, and hydrophobic contact area between the antigen and antibody.
[0140] After screening out the docking structure in which the interacting residues account for more than 0.5 of the CDR region, the embodiment of the present application also needs to evaluate at least one of the optimized screening indicators of the binding free energy of the antigen-antibody interface, the hydrophobic change before and after antigen-antibody binding, and the hydrophobic contact area between the antigen and antibody in the screened docking structure, so as to conduct further screening, so as to obtain a docking structure that is more stable and meets the binding change characteristics.
[0141] S37: Screen out the docking structure whose optimized screening index satisfies the preset screening conditions.
[0142] The preset screening conditions of the embodiment of the present application can be: the binding free energy of the antigen-antibody interface is less than -15 kcal / mol, the difference in SAP score before and after antigen-antibody binding is greater than 20, and the hydrophobic contact area between the antigen and antibody is greater than
[0143] The smaller the binding free energy of the antigen-antibody interface, the more spontaneous the antigen-antibody binding reaction is; the larger the difference in SAP scores before and after antigen-antibody binding and the hydrophobic contact area between the antigen and antibody, the greater the interaction after antigen-antibody binding and the higher the degree of complementarity.
[0144] That is to say, the second preset condition includes two conditions: the proportion mentioned in step S35 is greater than the fourth threshold and the optimization screening index mentioned in step S37 satisfies the preset screening condition.
[0145] In addition, S36 to S37 may also be executed before S33.
[0146] S40: Pre-mutate the residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure according to the frequencies of occurrence of different types of amino acids at different positions in the CDRH1 and CDRH2 regions of the natural antibody.
[0147] Since the antibody is a light chain antibody, only the CDR region of the antibody heavy chain in the docking structure was optimized.
[0148] And because the sequences of CDRH1 and CDRH2 regions are more regular and less diverse than CDRH3, the optimization of CDRH1 and CDRH2 regions can be performed by pre-mutating the residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure according to the frequencies of occurrence of different types of amino acids at different positions in the CDRH1 and CDRH2 regions in the natural antibody. For example, CDRH1 has 10 residues, and the frequencies of occurrence of different types of amino acids at each residue position are different, such as the frequency of occurrence of amino acid K at the first position is 20%, the frequency of occurrence of A is 15%, and the frequency of occurrence of amino acid T is 10%...; then when the residue at the first position in the CDRH1 region of the antibody in the docking structure is pre-mutated, since the probability of mutation to amino acid K, A, T... is related to the frequency of occurrence of these amino acids at the corresponding positions, the probability of mutation of the amino acid at the first position in the CDRH1 region of the antibody in the docking structure to amino acid K is 20%, the probability of mutation to amino acid A is 15%, and the probability of mutation to amino acid T is 10%, so that the residue mutations in the CDRH1 and CDRH2 regions of the antibody in the docking structure can be consistent with the frequency of occurrence of amino acids in natural antibodies.
[0149] Natural antibodies refer to various antibodies that exist naturally in the body of humans or animals without obvious infection or artificial injection of antigens, also known as normal antibodies. Pre-mutation can refer to virtual mutation, and the energy change is calculated through virtual mutation to determine whether to perform formal mutation.
[0150] S41: Determine whether to mutate or not the residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure based on the energy changes before and after the pre-mutation of the docking structure.
[0151] After the residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure are pre-mutated, the embodiment of the present application can determine the energy change before and after the pre-mutation of the docking structure by energy calculation. If the energy after the pre-mutation is lower than that before the pre-mutation, the residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure can be formally mutated according to the pre-mutation; if the energy after the pre-mutation is higher than that before the pre-mutation, then the formal mutation can be performed or not mutated according to a certain probability. Specifically, for example, a probability value P within 0 to 1 can be randomly generated each time the pre-mutation occurs, and the probability value P is compared with the preset value Po. If P is less than Po, a formal mutation is performed, otherwise no mutation is performed.
[0152] S42: pre-mutating the residues in the CDRH3 region of the antibody in the docking structure to preset residues other than cysteine and proline, and determining whether to mutate or not mutate the residues in the CDRH3 region of the antibody in the docking structure based on the energy change before and after the pre-mutation of the docking structure.
[0153] Since the sequence of the CDRH3 region is diverse and irregular, when optimizing the CDRH3 region, the residues in the CDRH3 region of the antibody in the docking structure can be mutated to preset residues except cysteine and proline, and the preset residues can be any other 17 residues different from the mutated residues.
[0154] Moreover, since cysteine is easily oxidized, which will affect the stability of the antibody, and proline is inaccurate in the calculation process, these two amino acids need to be excluded.
[0155] Furthermore, in the process of pre-mutating the residues in the CDRH3 region of the antibody in the docking structure, the embodiment of the present application can determine the energy change before and after the pre-mutation of the docking structure by energy calculation. If the energy after the pre-mutation is lower than that before the pre-mutation, the residues in the CDRH3 region of the antibody in the docking structure can be formally mutated according to the pre-mutation; if the energy after the pre-mutation is higher than that before the pre-mutation, then the formal mutation can be performed or not mutated according to a certain probability. Specifically, for example, a probability value P within 0 to 1 can be randomly generated each time a mutation occurs, and the probability value P is compared with a preset value Po. If P is less than Po, a formal mutation is performed, otherwise no mutation is performed.
[0156] S43: performing multiple iterative mutations on the residues in the CDRH1, CDRH2 and CDRH3 regions of the antibody in the docking structure in the manner of steps S40 to S42, and optimizing the structure of the binding interface of the docking structure after each mutation to obtain an optimized docking structure.
[0157] In the embodiment of the present application, the number of iterative mutations can be 100 times, 200 times or other empirical values, which are not limited in the present application. The docking structures obtained by mutating the residues in the CDRH1, CDRH2 and CDRH3 regions for the current time are all used as the structures for the next mutation. After each mutation, the binding interface of the docking structure is structurally optimized. After the iteration is completed, the docking structure after the last mutation and structural optimization is obtained as the optimized docking structure. After multiple iterative mutations and structural optimizations, such a docking structure can obtain an optimized docking structure.
[0158] S44: Repeat steps S40 to S43 multiple times to obtain multiple optimized docking structures and antibody sequences.
[0159] In the embodiment of the present application, each docking structure is subjected to a preset number of iterative mutations and structural optimization according to the above steps, and this process is repeated multiple times to obtain multiple optimized docking structures. For example, according to the above process, the calculation is preset 100 times, 200 times or other empirical values, so that each docking structure can generate 100 or 200 optimized docking structures.
[0160] S50: Calculate and obtain a first screening index of the plurality of optimized docking structures, wherein the first screening index comprises at least one of: binding free energy of the antigen-antibody interface, a hole score of the antigen-antibody interface, a change in hydrophobicity before and after antigen-antibody binding, and a hydrophobic contact area between the antigen and antibody.
[0161] In the first screening index of the embodiment of the present application, the smaller the binding free energy of the antigen-antibody interface and the pore score of the antigen-antibody interface, the better. The smaller the value, the tighter the binding of the antigen and antibody. The larger the change in hydrophobicity before and after antigen-antibody binding and the hydrophobic contact area between the antigen and antibody, the better. The larger the value, the greater the interaction after antigen-antibody binding and the higher the degree of complementarity.
[0162] S51: Screening out the optimized docking structure whose first screening index satisfies the first screening condition, and using the antibody in the optimized docking structure as the target antibody.
[0163] The first screening condition in the embodiment of the present application can be: the binding free energy of the antigen-antibody interface is less than -30 kcal / mol, the hole score of the antigen-antibody interface is less than 1, the difference in SAP score before and after antigen-antibody binding is greater than 26, and the hydrophobic contact area between the antigen and antibody is greater than
[0164] S52: Based on the structure of the target antigen and the sequence of the antibody in the optimized docking structure, the affinity of the target antigen and the antibody in the optimized docking structure is predicted using the affinity AI model, the optimized docking structure whose affinity meets the preset second screening condition is screened out, and the antibody in the optimized docking structure is used as the target antibody.
[0165] The second screening condition of the embodiment of the present application may be: sorting the affinities from high to low, and screening out the top 50% or other empirically optimized docking structures.
[0166] S53: Based on the sequence of the target antigen and the sequence of the antibody in the optimized docking structure, predict the second complex structure of the target antigen bound to the antibody in the optimized docking structure, as well as the non-binding structure of the antibody in the optimized docking structure.
[0167] In the embodiments of the present application, for example, Alphafold2 can be used to predict the complex structure of the target antigen bound to the antibody in the optimized docking structure, as well as the non-binding structure of the antibody in the optimized docking structure, that is, the separate structure of the antibody in the optimized docking structure.
[0168] S54: According to the second complex structure and the non-binding structure of the antibody in the optimized docking structure, a second screening index of the multiple optimized docking structures is calculated, wherein the second screening index includes at least one of the following indexes: confidence in the binding interface of the second complex structure; confidence in the relative position of the antigen and the antibody in the second complex structure; RMSD value of the antibody structure in the optimized docking structure and the antibody structure in the second complex structure after the structure of the antigen in the optimized docking structure and the structure of the antigen in the second complex structure are superimposed; RMSD value of the CDRH3 region of the antibody in the optimized docking structure and the CDRH3 region of the antibody in the second complex structure; RMSD value of the CDRH3 region of the non-binding structure of the antibody in the second complex structure and the optimized docking structure predicted by Alphafold2; overlap between the binding epitope of the antibody and the antigen in the second complex structure and the target epitope.
[0169] The confidence of the binding interface of the second complex structure indicates the credibility of the predicted binding interface. The higher the confidence, the higher the credibility of the structure and the higher the quality of the prediction.
[0170] The confidence level of the relative positions of the antigen and the antibody in the second complex structure indicates the credibility of the position where the antibody binds to the antigen. The higher the confidence level, the higher the credibility of the position where the antibody binds to the antigen.
[0171] After the structure of the antigen in the optimized docking structure is superimposed with the structure of the antigen in the second complex structure, the RMSD value of the antibody structure in the optimized docking structure and the antibody structure in the second complex structure indicates whether the binding posture of the antibody binding to the antigen in the optimized docking structure is consistent with that in the second complex structure predicted by Alphafold2. The second complex structure predicted by Alphafold2 is equivalent to the test. The smaller the RMSD, the more consistent the binding posture of the antibody binding to the antigen in the optimized docking structure is with that in the second complex structure predicted by Alphafold2, and the greater the possibility that this is the binding posture in the actual situation.
[0172] The RMSD value of the CDRH3 region of the antibody in the optimized docking structure and the CDRH3 region of the antibody in the second complex structure indicates the consistency of the structure of the CDRH3 region of the antibody in the optimized docking structure and the CDRH3 region of the antibody in the second complex structure predicted by Alphafold2. The CDRH3 region often plays a major role in the binding of antibodies to antigens, so the optimized docking structure is screened by the consistency of the CDRH3 structure. Here, the second complex structure predicted by Alphafold2 also acts as a test. The smaller the RMSD, the more consistent the structure of the CDRH3 region of the antibody in the optimized docking structure and the CDRH3 region of the antibody in the second complex structure, and the greater the possibility that the CDRH3 region presents such a structure in the real situation.
[0173] RMSD values of the CDRH3 region of the antibody structure in the second complex structure and the CDRH3 region of the non-binding structure of the antibody in the optimized docking structure predicted by Alphafold2: The second complex structure predicted by Alphafold2 and the non-binding structure of the antibody in the optimized docking structure predicted by Alphafold2 respectively represent the two structures of the antibody in the bound and non-bound states in the optimized docking structure. The CDRH3 region often plays a major role in the binding process. Therefore, the optimized docking structure is screened by the consistency of the CDRH3 region in the bound and non-binding states. The higher the consistency, the less significant conformational changes will occur in the CDRH3 region during the binding process, the higher the success rate of binding, and the greater the possibility of antigen-antibody binding in real conditions.
[0174] S55: Screening out the optimized docking structure whose second screening index satisfies the third screening condition, and using the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody.
[0175] The overlap between the binding epitope of the antibody and antigen in the second complex structure and the target epitope indicates the consistency between the computationally designed epitope and the target epitope. The smaller the deviation from the target epitope, the better. The antibody obtained in this way is the antibody that binds to the target epitope.
[0176] The third screening condition in the embodiment of the present application can be: the confidence of the binding interface of the second complex structure is greater than 80; the confidence of the relative position of the antigen and the antibody in the second complex structure is less than 20; after the structure of the antigen in the optimized docking structure is superimposed with the structure of the antigen in the second complex structure, the RMSD value of the antibody structure in the optimized docking structure and the antibody structure in the second complex structure is less than The RMSD value of the CDRH3 region of the antibody in the optimized docking structure and the CDRH3 region of the antibody in the second complex structure is less than The RMSD value of the CDRH3 region of the non-binding structure of the antibody structure in the second complex structure and the optimized docking structure predicted by Alphafold2 is less than The overlap between the binding epitopes of the antibody and antigen in the second complex structure and the target epitope is greater than 0.5.
[0177] In addition, the three calculation methods of S50 to S51, S52, and S53 to S55 can be combined to form a combined algorithm. These three calculation methods can be executed in the order of S50 to S51, S52, and S53 to S55, or in the order of S50 to S51, S53 to S55, and S52, or S52, S50 to S51, S53 to S55, or S52, S53 to S55, and S50 to S51, or S53 to S55, S50 to S51, and S52, or S53 to S55, S50 to S51, and S52, and the order of S53 to S55, S52, and S50 to S51 is not limited to this in this application.
[0178] The method for generating antibodies based on a human antibody framework library in the embodiment of the present application may have the following beneficial effects:
[0179] 1. The embodiments of the present application are based on a common light chain human antibody framework library, which can ensure that the designed target antibody is a human antibody.
[0180] 2. The embodiment of the present application first performs a low-precision screening of the human antibodies in the human antibody framework library, which can greatly reduce the subsequent calculation amount and improve the pass rate of the subsequent screening.
[0181] 3. The embodiments of the present application combine the human antibody backbone library and the docking method to achieve simultaneous docking and antibody side chain design, and obtain a stable docking structure that can bind to the target epitope region.
[0182] 4. After high-precision docking, restorative mutations are performed on the FR region residues of the antibody backbone to avoid affecting antibody expression and humanization.
[0183] 5. During the antibody optimization process, various stable optimizations were set for the CDRH3 region of the antibody to ensure the stability of the antibody, and the CDRH1 and CDRH2 regions of the antibody were optimized by amino acid frequency sampling to ensure the rationality of the optimization.
[0184] 6. In the process of combinatorial screening, a variety of scoring functions and models are used to screen antibodies from multiple angles, so as to improve the success rate of antibody design.
[0185] Through the above steps, the embodiments of the present application can efficiently and accurately design humanized antibodies targeting specific target epitopes from the human antibody skeleton library, providing strong support for the development of antibody drugs.
[0186] Corresponding to the aforementioned application function implementation method embodiment, the present application also provides a device, electronic device and corresponding embodiments for generating antibodies based on a human antibody framework library.
[0187] Figure 4 It is a schematic diagram of the structure of a device for generating antibodies based on a human antibody framework library shown in an embodiment of the present application.
[0188] See also Figure 4 The embodiment of the present application provides a device for generating antibodies based on a human antibody framework library, including an acquisition module 410, a prediction module 420, a docking module 430, an optimization module 440 and a screening module 450.
[0189] The acquisition module 410 is used to acquire the structure and sequence of the target antigen, the target epitope region, and the structure and sequence of the human antibody in the human antibody framework library.
[0190] The prediction module 420 is used to screen out human antibodies that meet the first preset condition from the human antibody framework library according to the sequence of the target antigen, the target epitope region, and the sequence of the human antibody in the human antibody framework library, and use them as candidate antibodies. Among them, the prediction module 420 is used to predict the first complex structure of the target antigen and the human antibody according to the sequence of the target antigen and the sequence of the human antibody; based on the binding interface and the target epitope region of the first complex structure, screen out the first complex structure that meets the first preset condition, and use the human antibody in the first complex structure that corresponds to the human antibody framework library as a candidate antibody.
[0191] The docking module 430 is used to dock the candidate antibody with the target antigen based on the target epitope region, adjust the docking posture of the candidate antibody, and screen the docking structure that meets the second preset condition; wherein the docking module 430 is used to screen out low-precision docking antigen residues and high-precision docking antigen residues from the residues of the target antigen structure according to the structure of the target antigen and the target epitope region; based on the low-precision docked antigen residues, the candidate antibody is docked with the target antigen with low precision, and the candidate antibody that meets shape complementarity and its relative position data with the target antigen are screened; based on the high-precision docked antigen residues and the candidate antibody that meets shape complementarity and its relative position data with the target antigen, the candidate antibody that meets shape complementarity is docked with the target antigen with high precision, and the docking posture of the candidate antibody that meets shape complementarity is adjusted during the high-precision docking process to obtain the docking structure.
[0192] The optimization module 440 is used to perform sequence optimization and structural optimization of the CDR region of the antibody in each docking structure based on a preset algorithm, and obtain multiple optimized docking structures and antibody sequences; wherein the optimization module 440 is used to pre-mutate the residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure according to the frequency of occurrence of different types of amino acids at different positions in the CDRH1 and CDRH2 regions in the natural antibody; determine whether the residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure are mutated or not according to the energy change before and after the pre-mutation of the docking structure; pre-mutate the residues in the CDRH3 region of the antibody in the docking structure to preset residues other than cysteine and proline, and determine whether the residues in the CDRH3 region of the antibody in the docking structure are mutated or not according to the energy change before and after the pre-mutation of the docking structure; perform multiple iterative mutations on the residues in the CDRH1, CDRH2 and CDRH3 regions of the antibody in the docking structure in the above manner, and perform structural optimization on the binding interface of the docking structure after each mutation to obtain an optimized docking structure; repeat the above process multiple times to obtain multiple optimized docking structures and antibody sequences.
[0193] The screening module 450 is used to perform multi-angle calculations on multiple optimized docking structures based on a preset combination algorithm, screen out the optimized docking structures that meet the preset combination screening conditions, and use the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody. Among them, the screening module 450 is used to: calculate and obtain a first screening index of multiple optimized docking structures, wherein the first screening index includes: at least one of the binding free energy of the antigen-antibody interface, the hole score of the antigen-antibody interface, the hydrophobic change before and after the antigen-antibody binding, and the hydrophobic contact area between the antigen and the antibody; screen out the optimized docking structure that meets the first screening condition, and use the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody; and / or, based on the structure of the target antigen and the sequence of the antibody in the optimized docking structure, predict the affinity of the target antigen and the antibody in the optimized docking structure, screen out the optimized docking structure whose affinity meets the preset second screening condition, and use the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody; and / or, based on the sequence of the target antigen and the sequence of the antibody in the optimized docking structure, predict the second complex structure of the target antigen and the antibody in the optimized docking structure, and the non-binding structure of the antibody in the optimized docking structure; according to the The second screening index of the multiple optimized docking structures is calculated based on the non-binding structure of the antibody in the two complex structures and the optimized docking structure, wherein the second screening index includes at least one of the following indexes: confidence of the binding interface of the second complex structure; confidence of the relative position of the antigen and the antibody in the second complex structure; RMSD value of the antibody structure in the optimized docking structure and the antibody structure in the second complex structure after the structure of the antigen in the optimized docking structure and the structure of the antigen in the second complex structure are superimposed; RMSD value of the CDRH3 region of the antibody in the optimized docking structure and the CDRH3 region of the antibody in the second complex structure; RMSD value of the CDRH3 region of the non-binding structure of the antibody structure in the second complex structure and the optimized docking structure; overlap between the binding epitope of the antibody and the antigen in the second complex structure and the target epitope; screen out the optimized docking structure whose second screening index meets the third screening condition, and use the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody.
[0194] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated again here.
[0195] Figure 5 It is a schematic diagram of the structure of an electronic device shown in an embodiment of the present application.
[0196] See also Figure 5, the electronic device 500 includes a memory 510 and a processor 520 .
[0197] 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, etc.
[0198] The memory 510 may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage devices. Among them, ROM can store static data or instructions required by the processor 520 or other modules of the computer. The permanent storage device may be a readable and writable storage device. The permanent storage device may be a non-volatile storage device that does not lose the stored instructions and data even after the computer is powered off. In some embodiments, the permanent storage device uses a large-capacity storage device (such as a magnetic or optical disk, flash memory) as a permanent storage device. In some other embodiments, the permanent storage device may be a removable storage device (such as a floppy disk, an optical drive). The system memory may be a readable and writable storage device or a volatile readable and writable storage device, such as a dynamic random access memory. The system memory may store some or all instructions and data required by the processor at run time. In addition, the memory 510 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (such as DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and 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 laser disc (CD), a read-only digital versatile disc (such as a DVD-ROM, a double-layer DVD-ROM), a read-only Blu-ray disc, an ultra-density optical disc, a flash memory card (such as an SD card, a mini SD card, a Micro-SD card, etc.), a magnetic floppy disk, etc. The computer-readable storage medium does not include carrier waves and transient electronic signals transmitted wirelessly or wired.
[0199] The memory 510 stores executable codes, and when the executable codes are processed by the processor 520 , the processor 520 can execute part or all of the methods described above.
[0200] 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.
[0201] 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.
[0202] The embodiments of the present application 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 improvements to the technology 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 generating antibodies based on a human antibody framework library, characterized in that: include: Obtain the structure and sequence of the target antigen, the target epitope region, and the structure and sequence of the human antibodies in the human antibody framework library; According to the sequence of the target antigen, the target epitope region, and the sequence of the human antibodies in the human antibody framework library, human antibodies that meet the first preset condition are screened from the human antibody framework library as candidate antibodies; Based on the target epitope region, docking the candidate antibody with the target antigen, adjusting the docking posture of the candidate antibody, and screening to obtain a docking structure that meets a second preset condition; Based on a preset algorithm, sequence optimization is performed on the CDR region of the antibody in each docking structure and structure optimization is performed on the binding interface to obtain multiple optimized docking structures and antibody sequences; Based on a preset combination algorithm, a plurality of the optimized docking structures and antibody sequences are calculated from multiple angles to screen out the optimized docking structures and antibody sequences that meet the preset combination screening conditions, and the structure and sequence of the antibody in the optimized docking structure are used as the structure and sequence of the target antibody.
2. The method according to claim 1, characterized in that The step of screening human antibodies that meet the first preset condition from the human antibody framework library according to the sequence of the target antigen, the target epitope region, and the sequence of the human antibodies in the human antibody framework library as candidate antibodies comprises: Predicting a first complex structure of the target antigen and the human antibody based on the sequence of the target antigen and the sequence of the human antibody; According to the binding interface of the first complex structure and the target epitope region, a first complex structure that meets the first preset condition is screened, and the human antibodies in the first complex structure corresponding to the human antibody framework library are taken as candidate antibodies.
3. The method according to claim 1, characterized in that The step of docking the candidate antibody with the target antigen based on the target epitope region, adjusting the docking posture of the candidate antibody, and screening to obtain a docking structure that meets the second preset condition includes: According to the structure of the target antigen and the target epitope region, low-precision docking antigen residues and high-precision docking antigen residues are screened from the residues of the target antigen structure; Based on the low-precision docked antigen residues, the candidate antibody is docked with the target antigen with low precision, and candidate antibodies that satisfy shape complementarity and their relative position data with the target antigen are screened and obtained; Based on the antigen residues for high-precision docking, the candidate antibody satisfying shape complementarity and its relative position data with the target antigen, the candidate antibody satisfying shape complementarity is docked with the target antigen with high precision. During the high-precision docking process, the docking posture of the candidate antibody satisfying shape complementarity is adjusted to obtain a docking structure.
4. The method according to claim 3, characterized in that: The step of screening low-precision docking antigen residues and high-precision docking antigen residues from the residues of the target antigen structure according to the structure of the target antigen and the target epitope region comprises: Analyzing the hydrophobicity of all residues in the structure of the target antigen, and screening out residues with a hydrophobicity greater than a third threshold as candidate residues; Selecting clustered residues from the residues contained in the target epitope region, and selecting a first preset number of residues from the clustered residues with the candidate residues as the preferred candidate residues as antigen residues for low-precision docking; The low-precision docked antigen residues and residues in their neighborhood form a second preset number of high-precision docked antigen residues, wherein the residues in the neighborhood of the low-precision docked antigen residues belong to the residues contained in the target epitope region; and / or, The method of docking the candidate antibody with the target antigen based on the low-precision docked antigen residues to obtain candidate antibodies that meet shape complementarity and their relative position data with the target antigen includes: mutating all the residues on the main chain of the candidate antibody to valine; Adjusting the posture of the candidate antibody, screening out the candidate antibody whose main chain can be shape-complementary with the target epitope region of the target antigen, and obtaining the candidate antibody satisfying shape complementarity and its relative position data with the target antigen; and / or, Based on the antigen residues for high-precision docking and the candidate antibody satisfying shape complementarity and the relative position data between the candidate antibody satisfying shape complementarity and the target antigen, the candidate antibody satisfying shape complementarity is docked with the target antigen with high precision, and during the high-precision docking process, the docking posture of the candidate antibody satisfying shape complementarity is adjusted to obtain a docking structure, including: Based on the antigen residue docked with high precision, generating a preset side chain that interacts with the antigen residue near the designated antigen residue docked with high precision; Calculating the binding energy data of the preset side chain and the target antigen, and storing the preset side chain and the corresponding binding energy data to obtain a side chain matching library; Based on the candidate antibody satisfying shape complementarity and the relative position data between the candidate antibody satisfying shape complementarity and the target antigen, the docking posture of the candidate antibody satisfying shape complementarity is adjusted, and the side chain of the candidate antibody satisfying shape complementarity is matched with the preset side chain in the side chain matching library to obtain the binding energy data corresponding to the side chain in the candidate antibody satisfying shape complementarity that matches the preset side chain in the side chain matching library; According to the binding energy data of the side chains in the candidate antibodies satisfying shape complementarity, the side chains in the candidate antibodies satisfying shape complementarity are replaced with preset side chains or retained to obtain a docking structure.
5. The method according to claim 3, characterized in that: After the docking structure is obtained, the method further comprises: Comparing the FR region of the candidate antibody and the antibody in the docking structure, mutating the residues in the FR region of the antibody in the docking structure to the same residues as those in the FR region of the candidate antibody; and / or, further comprising the step of screening the docking structure that meets the second preset condition: analyzing the interacting residues on the antibody in the docking structure and the target antigen; Calculating the proportion of the interacting residues located in the CDR region of the antibody in the docking structure, and screening out docking structures having a proportion greater than a fourth threshold; Calculating the optimization screening index of the docking structure after screening, wherein the optimization screening index includes at least one of: binding free energy of the antigen-antibody interface, hydrophobic change before and after antigen-antibody binding, and hydrophobic contact area between antigen and antibody; The docking structure whose optimized screening index meets the preset screening condition is screened out.
6. The method according to claim 1, characterized in that Based on the preset algorithm, the CDR region of the antibody in each docking structure is sequence optimized and the binding interface is structurally optimized to obtain multiple optimized docking structures and antibody sequences, including: Pre-mutating the residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure according to the frequencies of occurrence of different types of amino acids at different positions in the CDRH1 and CDRH2 regions of the natural antibody; Determining whether to mutate or not mutate residues in the CDRH1 and CDRH2 regions of the antibody in the docking structure according to energy changes before and after the pre-mutation of the docking structure; Premutating the residues in the CDRH3 region of the antibody in the docking structure to preset residues other than cysteine and proline, and determining whether to mutate or not mutate the residues in the CDRH3 region of the antibody in the docking structure according to the energy change before and after the premutation of the docking structure; According to the above method, the residues in the CDRH1, CDRH2 and CDRH3 regions of the antibody in the docking structure are iteratively mutated multiple times, and the binding interface of the docking structure after each mutation is structurally optimized to obtain an optimized docking structure; The above process was repeated multiple times to obtain multiple optimized docking structures and antibody sequences.
7. The method according to claim 1, characterized in that The method of performing multi-angle calculations on the plurality of optimized docking structures and antibody sequences based on a preset combination algorithm, screening out the optimized docking structures and antibody sequences that meet the preset combination screening conditions, and using the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody, comprises: Calculating and obtaining a plurality of first screening indicators of the optimized docking structures, wherein the first screening indicator comprises at least one of: binding free energy of the antigen-antibody interface, a hole score of the antigen-antibody interface, a change in hydrophobicity before and after antigen-antibody binding, and a hydrophobic contact area between the antigen and antibody; Screening out the optimized docking structure whose first screening index satisfies the first screening condition, and using the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody; and / or, According to the structure of the target antigen and the sequence of the antibody in the optimized docking structure, the affinity of the target antigen and the antibody in the optimized docking structure is predicted, the optimized docking structure whose affinity meets the preset second screening condition is screened out, and the structure and sequence of the antibody in the optimized docking structure are used as the structure and sequence of the target antibody; and / or, According to the sequence of the target antigen and the sequence of the antibody in the optimized docking structure, predicting a second complex structure of the target antigen bound to the antibody in the optimized docking structure, and a non-binding structure of the antibody in the optimized docking structure; According to the second complex structure and the non-binding structure of the antibody in the optimized docking structure, a plurality of second screening indicators of the optimized docking structure are calculated, wherein the second screening indicator includes at least one of the following indicators: Confidence of the binding interface of the second complex structure; Confidence of the relative positions of the antigen and the antibody in the second complex structure; RMSD value of the antibody structure in the optimized docking structure and the antibody structure in the second complex structure after the structure of the antigen in the optimized docking structure is superimposed with the structure of the antigen in the second complex structure; RMSD value of the CDRH3 region of the antibody in the optimized docking structure and the CDRH3 region of the antibody in the second complex structure; RMSD value of the CDRH3 region of the antibody structure in the second complex structure and the non-binding structure of the antibody in the optimized docking structure; overlap between the binding epitope of the antibody and the antigen in the second complex structure and the target epitope; The optimized docking structure whose second screening index satisfies the third screening condition is screened out, and the structure and sequence of the antibody in the optimized docking structure are used as the structure and sequence of the target antibody.
8. A device for generating antibodies based on a human antibody framework library, characterized in that: include: An acquisition module, used to obtain the structure and sequence of the target antigen, the target epitope region, and the structure and sequence of the human antibody in the human antibody framework library; A prediction module, for screening human antibodies that meet a first preset condition from the human antibody framework library as candidate antibodies based on the sequence of the target antigen, the target epitope region, and the sequence of the human antibodies in the human antibody framework library; A docking module, used for docking the candidate antibody with the target antigen based on the target epitope region, adjusting the docking posture of the candidate antibody, and screening to obtain a docking structure that meets a second preset condition; An optimization module, for performing sequence optimization on the CDR region of the antibody in each docking structure and structural optimization on the binding interface based on a preset algorithm, to obtain a plurality of optimized docking structures and antibody sequences; A screening module is used to perform multi-angle calculations on the plurality of optimized docking structures based on a preset combination algorithm, screen out the optimized docking structures and sequences that meet the preset combination screening conditions, and use the structure and sequence of the antibody in the optimized docking structure as the structure and sequence of the target antibody.
9. 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 7.
10. 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 7.
Citation Information
Cited By
Targeted target epitope antibody design method
CN120833853A
A method for antibody design targeting a target epitope
CN120833853B
Iterative nearest point-based antigen-antibody binding conformation optimization method and device
CN120913633A
