Method, device, electronic device and storage medium for determining antibody sequence

Abstract

The present disclosure provides a method, device, electronic device and storage medium for determining an antibody sequence, which relates to the field of artificial intelligence technology, and specifically to the field of biocomputing technology. The specific implementation scheme is as follows: starting from a reference antibody sequence of an antigen sequence, the reference antibody sequence is iteratively optimized; for the i+1th iteration, a complex conformation i of the antigen sequence and the antibody sequence i obtained in the i-th iteration is obtained, where i is a natural number greater than or equal to 1; based on the complex conformation i, the antibody properties of the antibody sequence i are obtained, and the antibody properties of the antibody sequence i are expected to be verified; if the antibody properties of the antibody sequence i pass the expected verification, the antibody sequence i is determined to be the target antibody sequence of the antigen sequence; if the antibody properties of the antibody sequence i do not pass the expected verification, the antibody sequence i+1 is determined based on the reference antibody sequence, and the i+2th iteration is continued.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, specifically to the field of biological computing technology, and in particular to a method, apparatus, electronic device, and storage medium for determining antibody sequences. Background Technology

[0002] Macromolecular drugs possess high targeting and specificity, and their design aims to develop biomolecules with therapeutic potential, such as proteins, antibodies, and nucleic acids. However, current antibody drug design relies on wet laboratory experiments, which are costly and inefficient. Summary of the Invention

[0003] This disclosure provides a method, apparatus, electronic device, and storage medium for determining antibody sequences.

[0004] According to one aspect of this disclosure, a method for determining an antibody sequence is provided, comprising: starting from a reference antibody sequence of an antigen sequence, iteratively optimizing the reference antibody sequence; for the (i+1)th iteration, obtaining a complex conformation i of the antigen sequence and the antibody sequence i obtained in the i-th iteration, wherein i is a natural number greater than or equal to 1; obtaining the antibody properties of the antibody sequence i based on the complex conformation i, and performing expected verification on the antibody properties of the antibody sequence i; if the antibody properties of the antibody sequence i pass the expected verification, determining the antibody sequence i as the target antibody sequence of the antigen sequence; if the antibody properties of the antibody sequence i fail the expected verification, determining antibody sequence i+1 based on the reference antibody sequence, and continuing to execute the (i+2)th iteration.

[0005] According to another aspect of this disclosure, an apparatus for determining an antibody sequence is provided, comprising: a first iterative optimization module, configured to iteratively optimize a reference antibody sequence starting from an antigen sequence; an acquisition module, configured to acquire, for the (i+1)th iteration, a complex conformation i of the antigen sequence and the antibody sequence i obtained in the i-th iteration, wherein i is a natural number greater than or equal to 1; a verification module, configured to acquire the antibody properties of the antibody sequence i based on the complex conformation i, and perform expected verification on the antibody properties of the antibody sequence i; an output module, configured to determine the antibody sequence i as the target antibody sequence of the antigen sequence if the antibody properties of the antibody sequence i pass the expected verification; and a second iterative optimization module, configured to determine an antibody sequence i+1 based on the reference antibody sequence if the antibody properties of the antibody sequence i fail the expected verification, and continue to execute the (i+2)th iteration.

[0006] According to another aspect of this disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the antibody sequence determination method described in one aspect of the above embodiments.

[0007] According to another aspect of this disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are stored thereon for causing the computer to perform the antibody sequence determination method described in one aspect of the above-described embodiment.

[0008] According to another aspect of this disclosure, a computer program product is provided, including a computer program / instructions that, when executed by a processor, implement the antibody sequence determination method described in one aspect of the embodiments above.

[0009] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0010] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0011] Figure 1 A flowchart illustrating a method for determining an antibody sequence provided in this embodiment of the disclosure;

[0012] Figure 2 A flowchart illustrating another method for determining an antibody sequence provided in this embodiment of the disclosure;

[0013] Figure 3 A flowchart illustrating another method for determining an antibody sequence provided in this embodiment of the disclosure;

[0014] Figure 4 A flowchart illustrating another method for determining an antibody sequence provided in this embodiment of the disclosure;

[0015] Figure 5 A flowchart illustrating another method for determining an antibody sequence provided in this embodiment of the disclosure;

[0016] Figure 6 This is a schematic diagram of the antibody sequence optimization process provided in the embodiments of this disclosure;

[0017] Figure 7 A schematic diagram of an antibody sequence determination device provided in an embodiment of this disclosure;

[0018] Figure 8 This is a block diagram of an electronic device used to implement the antibody sequence determination method of the embodiments of this disclosure. Detailed Implementation

[0019] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0020] The following description, with reference to the accompanying drawings, outlines a method, apparatus, and electronic device for determining antibody sequences according to embodiments of the present disclosure.

[0021] Artificial intelligence (AI) is the study of enabling computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It involves both hardware and software technologies. AI hardware technologies generally include computer vision, speech recognition, natural language processing, and related technologies such as deep learning, big data processing, and knowledge graphs.

[0022] Biological computing is a field that draws on the principles and mechanisms of biological systems to solve computational problems. It applies biological characteristics and processes to computational systems to improve computational efficiency and performance. The goal of biological computing is to draw inspiration from biological systems and translate it into new computational methods and technologies to solve complex problems. It has wide applications in optimization, pattern recognition, data analysis, and simulation, and is continuously developing and expanding.

[0023] The antibody sequence determination method provided in this disclosure can be applied to fields such as antibody drug design, chimeric antigen receptor T-cell immunotherapy (CAR-T), chimeric antigen receptor NK cell therapy (CAR-NK), and antibody-drug conjugates (ADC) design.

[0024] Figure 1 This is a flowchart illustrating a method for determining an antibody sequence provided in an embodiment of this disclosure.

[0025] like Figure 1 As shown, the method for determining the antibody sequence may include:

[0026] S101, starting with the reference antibody sequence of the antigen sequence, iteratively optimizes the reference antibody sequence.

[0027] It should be noted that the execution entity of the antibody sequence determination method in this disclosure embodiment can be a hardware device with data processing capabilities and / or the necessary software to drive the hardware device. Optionally, the execution entity may include a server, a user terminal, and other smart devices. Optionally, the user terminal includes, but is not limited to, mobile phones, computers, smart voice interaction devices, etc. Optionally, the server includes, but is not limited to, a network server, an application server, or a server of a distributed system, or a server combined with blockchain, etc. This disclosure embodiment does not impose specific limitations.

[0028] Understandably, an antigen sequence refers to the sequence of a biomolecule (such as a protein) or chemical substance that can be recognized by the immune system and trigger an immune response. For example, in immunology, an antigen sequence typically refers to the amino acid sequence of a protein antigen.

[0029] An antibody sequence is a sequence encoded by immunoglobulin genes in the genome. In humans, immunoglobulin genes have multiple alleles, and different types of antibodies are produced through gene rearrangement and mutation. Each antibody has its unique amino acid sequence that binds to a specific antigen.

[0030] In some implementations, antigen sequences and reference antibody sequences can be obtained based on sequencing technologies and genomic methods, and / or antigen sequences and reference antibody sequences can be obtained from relevant databases.

[0031] Optionally, by predicting the complex conformation and antibody properties of the antigen sequence and the reference antibody sequence, an antibody sequence that meets the expectations is obtained and used as the optimized antibody sequence. Simultaneously, the antibody sequences that do not meet the expectations are subjected to the next iteration of optimization, thus realizing the iterative optimization process of the reference antibody sequence.

[0032] S102, for the (i+1)th iteration, obtain the complex conformation i of the antigen sequence and the antibody sequence i obtained in the i-th iteration, where i is a natural number greater than or equal to 1.

[0033] In some implementations, artificial intelligence can be used to predict the complex conformation of antigen and antibody sequences, thus obtaining the complex conformation. Optionally, a predictive model for the complex conformation can be pre-trained, and the antigen and antibody sequences can be input into the model, which then outputs the complex conformation.

[0034] Optionally, for the (i+1)th iteration, the antigen sequence and the antibody sequence i generated in the i-th iteration are input into a pre-trained prediction model of the complex conformation, and the model outputs the complex conformation i of the (i+1)th iteration.

[0035] S103. Based on the complex conformation i, obtain the antibody properties of antibody sequence i, and perform expected verification of the antibody properties of antibody sequence i.

[0036] In some implementations, antibody properties include, but are not limited to, antibody affinity to antigen, antibody immunogenicity, and antibody solubility. Optionally, antibody properties can be predicted for complex conformation i separately based on multiple pre-trained antibody property prediction models to obtain multiple antibody properties of antibody sequence i. Alternatively, antibody properties can be predicted for complex conformation i based on a single antibody property prediction model, and multiple antibody properties of antibody sequence i can be jointly output.

[0037] For example, the affinity of antibody sequence i can be predicted based on molecular dynamics (Molecular Mechanics Generalized Born Surface Area, MMGBSA). The immunogenicity of antibody sequence i can be predicted using a pre-trained property prediction model.

[0038] Furthermore, after obtaining the antibody properties of antibody sequence i, wet experimental methods can be used to verify the desired antibody properties of antibody sequence i, and highly accurate molecular dynamics methods can also be used to verify the desired antibody properties of antibody sequence i. For example, surface plasmon resonance (SPR) experiments can be used to verify the desired antigen-antibody affinity.

[0039] S104. If the antibody properties of antibody sequence i pass the expected verification, antibody sequence i is determined to be the target antibody sequence of the antigen sequence.

[0040] In some implementations, the magnitude of the antibody properties of antibody sequence i can be used to determine whether the antibody properties of antibody sequence i pass the expected verification. Alternatively, a threshold for antibody properties can be used to determine the conditions for passing the expected verification, and then the antibody properties of antibody sequence i can be used to determine whether it passes the expected verification.

[0041] For example, let antibody properties include the affinity between the antibody and the antigen and antibody immunogenicity, with the threshold corresponding to affinity being called the affinity threshold and the threshold corresponding to antibody immunogenicity being called the immunogenicity threshold. Then, the conditions for antibody properties of antibody sequence i to pass the expected verification are as follows: the affinity property value of antibody sequence i is greater than the affinity threshold, and the immunogenicity property value of antibody sequence i is less than the immunogenicity threshold.

[0042] Optionally, in response to the affinity attribute value of antibody sequence i being greater than the affinity threshold and the immunogenicity attribute value of antibody sequence i being less than the immunogenicity threshold, if the antibody attribute of antibody sequence i is determined to pass the expected verification, then antibody sequence i is determined to be the target antibody sequence of the antigen sequence.

[0043] S105, if the antibody properties of antibody sequence i fail the expected verification, determine antibody sequence i+1 based on the reference antibody sequence, and continue to execute the i+2th iteration.

[0044] In some implementations, when the antibody properties of antibody sequence i fail the expected verification, one or more candidate antibody sequences can be obtained by mutating the reference antibody sequence. Based on the antibody properties of the candidate antibody sequences, one or more antibody sequences can be selected from the candidate antibody sequences as antibody sequence i+1, and the i+2th iteration can be performed on antibody sequence i+1.

[0045] Optionally, candidate antibody sequences can be sorted based on their antibody properties, and the candidate antibody sequence with the highest sorting result can be selected as antibody sequence i+1.

[0046] According to the antibody sequence determination method provided in this disclosure, by obtaining a reference antibody sequence for the antigen sequence and iteratively optimizing the reference antibody sequence, antibody R&D costs can be reduced and the success rate of antibody design can be improved. Based on a prediction model, the complex conformation of the antigen and antibody sequences is predicted, and then the antibody properties of the antibody sequence are predicted based on the complex conformation, so as to perform expected verification of antibody properties and achieve simultaneous optimization of multiple antibody properties. When the antibody properties pass the expected verification, the target antibody sequence is obtained; when the antibody properties fail the expected verification, the antibody sequence continues to be iteratively optimized. Through iterative iteration, the efficiency and success rate of antibody design can be improved. By using conformation prediction and property prediction, targeted design optimization of antibodies can be achieved, thereby improving the design success rate and saving a large amount of wet experimental processes, thus reducing design costs.

[0047] Figure 2 This is a flowchart illustrating a method for determining an antibody sequence provided in an embodiment of this disclosure.

[0048] like Figure 2 As shown, the method for determining the antibody sequence may include:

[0049] S201, starting with the reference antibody sequence of the antigen sequence, iteratively optimizes the reference antibody sequence.

[0050] The details of step S201 can be found in the above embodiments and will not be repeated here.

[0051] S202, for the (i+1)th iteration, the antigen sequence and antibody sequence i are input into the pre-trained conformation prediction model, and the conformation prediction model performs conformation prediction to obtain the complex conformation i.

[0052] In some implementations, a conformation prediction model can be pre-trained to predict the conformation of the complex. The antigen and antibody sequences are then input into the model, which outputs the conformation of the complex. Predicting the conformation of the complex can help understand the interaction mechanisms and structural characteristics between antigens and antibodies, thus aiding in drug design and optimization.

[0053] Optionally, for the (i+1)th iteration, the complex conformation i can be obtained by inputting the antigen sequence and antibody sequence i into a pre-trained conformation prediction model, which then predicts the complex conformation. For example, the antigen sequence and antibody sequence i can be input into an AI-based conformation prediction tool to obtain the complex conformation i.

[0054] S203, input the complex conformation i into the prediction tools corresponding to various antibody properties respectively, and use the prediction tools of various antibody properties to predict the properties of the complex conformation i to obtain the antibody properties of antibody sequence i.

[0055] In some implementations, to improve the accuracy and effectiveness of predicting antibody properties, different prediction tools can be used to predict different antibody properties, ultimately yielding antibody sequences.

[0056] Optionally, the antibody properties of antibody sequence i include at least the antibody affinity to the antigen and antibody immunogenicity. The affinity of antibody sequence i can be predicted based on molecular dynamics (Molecular Mechanics Generalized Born Surface Area, MMGBSA). The antibody immunogenicity of antibody sequence i can be predicted using a pre-trained property prediction model.

[0057] In some implementations, a multifunctional prediction joint model can also be used to predict the properties of complex conformation i. The prediction joint model includes different network branches to predict different antibody properties. It can combine the advantages of multiple prediction tools, comprehensively consider the correlation and interaction between different antibody properties, and provide more comprehensive and accurate prediction results.

[0058] Optionally, by calling the antibody property prediction joint model, the complex conformation i is input into different prediction network branches in the antibody property prediction joint model, where different prediction network branches are applied to predict different antibody properties. Each prediction network branch performs property prediction on the complex conformation i to obtain the antibody properties of the antibody sequence i.

[0059] For example, suppose the antibody property prediction joint model includes network branch A and network branch B, where network branch A is used to predict the affinity between the antibody and the antigen, and network branch B is used to predict the immunogenicity of the antibody. When complex conformation i is input into the antibody property prediction joint model, network branch A outputs the affinity between the antibody and the antigen for complex conformation i, and network branch B outputs the immunogenicity of the antibody for complex conformation i.

[0060] S204, to perform expected verification of the antibody properties of antibody sequence i.

[0061] S205, if the antibody properties of antibody sequence i pass the expected verification, antibody sequence i is determined to be the target antibody sequence of the antigen sequence.

[0062] S206, if the antibody properties of antibody sequence i fail the expected verification, determine antibody sequence i+1 based on the reference antibody sequence, and continue to execute the i+2th iteration.

[0063] The details of steps S204-S206 can be found in the above embodiments and will not be repeated here.

[0064] According to the antibody sequence determination method provided in this disclosure, by obtaining a reference antibody sequence for the antigen sequence and iteratively optimizing the reference antibody sequence, antibody R&D costs can be reduced and the success rate of antibody design can be improved. Based on a prediction model, the complex conformation of the antigen and antibody sequences is predicted, and then the antibody properties of the antibody sequence are predicted based on the complex conformation, so as to perform expected verification of antibody properties and achieve simultaneous optimization of multiple antibody properties. Different prediction tools can be used to predict various antibody properties to improve the accuracy of prediction, and a joint prediction model can also be used to predict antibody properties, providing comprehensive prediction results. When the antibody properties pass the expected verification, the target antibody sequence is obtained; when the antibody properties fail the expected verification, the antibody sequence continues to be iteratively optimized. Through iterative iteration, the efficiency and success rate of antibody design can be improved. By using conformation prediction and property prediction, targeted design optimization of antibodies can be achieved, thereby improving the design success rate and saving a large amount of wet experimental processes, thus saving design costs.

[0065] Figure 3 This is a flowchart illustrating a method for determining an antibody sequence provided in an embodiment of this disclosure.

[0066] like Figure 3 As shown, the method for determining the antibody sequence may include:

[0067] S301, starting with the reference antibody sequence of the antigen sequence, iteratively optimizes the reference antibody sequence.

[0068] S302, for the (i+1)th iteration, obtain the complex conformation i of the antigen sequence and the antibody sequence i obtained in the i-th iteration, where i is a natural number greater than or equal to 1.

[0069] S303, based on complex conformation i, obtain the antibody properties of antibody sequence i.

[0070] The relevant content of steps S301-S303 can be found in the above embodiments, and will not be repeated here.

[0071] S304, obtain the determination criteria for each type of antibody attribute of antibody sequence i.

[0072] In some implementations, the antibody properties of antibody sequence i include at least antibody affinity to the antigen and antibody immunogenicity. The criteria for determining each type of antibody property of antibody sequence i can be determined based on thresholds for antibody affinity to the antigen and antibody immunogenicity.

[0073] Optionally, the criteria for determining the affinity between an antibody and an antigen may be that the affinity of antibody sequence i is greater than the affinity threshold; the criteria for determining the immunogenicity of an antibody may be that the immunogenicity of antibody sequence i is less than the immunogenicity threshold.

[0074] S305, compare the attribute value of each type of antibody attribute of antibody sequence i with its respective judgment condition to verify the antibody attribute of antibody sequence i.

[0075] In some implementations, by obtaining each type of antibody attribute of antibody sequence i and determining the attribute value of each type of antibody attribute, the antibody attributes of antibody sequence i can be verified by comparing the attribute value with its corresponding judgment condition, so as to verify whether the antibody attributes pass the expected verification.

[0076] Optionally, in response to the affinity attribute value of antibody sequence i being greater than the affinity threshold and the antibody immunogenicity attribute value of antibody sequence i being less than the antibody immunogenicity threshold, the antibody attribute is determined to have passed the expected verification; otherwise, the antibody attribute fails the expected verification.

[0077] S306, If the antibody properties of antibody sequence i pass the expected verification, antibody sequence i is determined to be the target antibody sequence of the antigen sequence.

[0078] S307, if the antibody properties of antibody sequence i fail the expected verification, determine antibody sequence i+1 based on the reference antibody sequence, and continue to execute the i+2th iteration.

[0079] The details of steps S306-S307 can be found in the above embodiments and will not be repeated here.

[0080] In some implementations, if the antibody properties of antibody sequence i are not verified, the antibody property prediction joint model can be optimized and fine-tuned based on the property information of antibody sequence i and the verification result data to improve the prediction accuracy of antibody properties.

[0081] According to the antibody sequence determination method provided in this disclosure, by obtaining a reference antibody sequence for the antigen sequence and iteratively optimizing the reference antibody sequence, antibody R&D costs can be reduced and the success rate of antibody design can be improved. Based on a prediction model, the complex conformation of the antigen and antibody sequences is predicted, and then the antibody properties of the antibody sequence are predicted based on the complex conformation. Based on the attribute thresholds for each type of antibody property, judgment conditions are determined, and the antibody properties are then validated according to expectations. This allows for simultaneous optimization of multiple antibody properties. When the antibody properties pass the expectation validation, the target antibody sequence is obtained. When the antibody properties fail the expectation validation, the antibody sequence continues to be iteratively optimized. Through iterative iteration, the efficiency and success rate of antibody design can be improved. By using conformation prediction and property prediction, targeted design optimization of antibodies can be achieved, thereby improving the design success rate and eliminating a large amount of wet experimental processes, thus saving design costs.

[0082] Figure 4 This is a flowchart illustrating a method for determining an antibody sequence provided in an embodiment of this disclosure.

[0083] like Figure 4 As shown, the method for determining the antibody sequence may include:

[0084] S401 starts with the reference antibody sequence of the antigen sequence and iteratively optimizes the reference antibody sequence.

[0085] S402, for the (i+1)th iteration, obtain the complex conformation i of the antigen sequence and the antibody sequence i obtained in the i-th iteration, where i is a natural number greater than or equal to 1.

[0086] S403, based on the complex conformation i, obtain the antibody properties of antibody sequence i, and perform expected verification of the antibody properties of antibody sequence i.

[0087] S404, If the antibody properties of antibody sequence i pass the expected verification, antibody sequence i is determined to be the target antibody sequence of the antigen sequence.

[0088] The relevant content of steps S401-S404 can be found in the above embodiments, and will not be repeated here.

[0089] S405, if the antibody properties of antibody sequence i fail the expected verification, generate at least one candidate antibody sequence based on the reference antibody sequence.

[0090] In some implementations, when the antibody properties of antibody sequence i fail to meet the expected validation, the performance of the antibody sequence can be improved by mutating the reference antibody sequence, thereby optimizing the antibody properties of the antibody sequence.

[0091] Optionally, at least one candidate antibody sequence can be obtained by acquiring the mutation sites of a reference antibody sequence and mutating the reference antibody sequence based on the mutation sites. If the reference antibody sequence has one mutation site, a candidate antibody sequence can be obtained by mutating the mutation site. If the reference antibody sequence has multiple mutation sites, multiple candidate antibody sequences can be obtained by mutating the mutation sites.

[0092] S406, Select one or more antibody sequences from at least one candidate antibody sequence as antibody sequence i+1.

[0093] In some implementations, one or more antibody sequences can be selected from candidate antibody sequences based on the distribution of antibody properties, and designated as antibody sequence i+1. Alternatively, the antibody sequence with the best antibody properties can be selected from the candidate antibody sequences as antibody sequence i+1, thereby optimizing the reference antibody sequence.

[0094] Optionally, the attribute probability distributions of candidate antibody sequences and antigen sequences can be obtained based on the attribute information of antibody sequence i, and antibody sequence i+1 can be selected from at least one candidate antibody sequence based on the attribute probability distribution of candidate antibody sequences.

[0095] Optionally, candidate antibody sequences can be sorted based on the probability distribution of attributes, and the top-ranked candidate antibody sequence can be selected as antibody sequence i+1. In other words, candidate antibody sequences can be sampled based on the probability distribution of attributes to obtain one or more top-ranked antibody sequences as antibody sequence i+1.

[0096] S407, continue with the (i+2)th iteration.

[0097] The details of step S407 can be found in the above embodiments and will not be repeated here.

[0098] According to the antibody sequence determination method provided in this disclosure, by obtaining a reference antibody sequence for the antigen sequence and iteratively optimizing the reference antibody sequence, antibody R&D costs can be reduced and the success rate of antibody design can be improved. Based on a prediction model, the complex conformation of the antigen and antibody sequences is predicted, and then the antibody properties of the antibody sequence are predicted based on the complex conformation, so as to perform expected verification of antibody properties and achieve simultaneous optimization of multiple antibody properties. When the antibody properties pass the expected verification, the target antibody sequence is obtained; when the antibody properties fail the expected verification, the reference antibody sequence is mutated to obtain candidate antibody sequences, and antibody sequences are selected from the candidate antibody sequences for continued iterative optimization. Through iterative iteration, the efficiency and success rate of antibody design can be improved. By using conformation prediction and property prediction, targeted design and optimization of antibodies can be achieved, thereby improving the design success rate and saving a large amount of wet experimental processes, thus reducing design costs.

[0099] Figure 5 This is a flowchart illustrating a method for determining an antibody sequence provided in an embodiment of this disclosure.

[0100] like Figure 5 As shown, the method for determining the antibody sequence may include:

[0101] S501 starts with the reference antibody sequence of the antigen sequence and iteratively optimizes the reference antibody sequence.

[0102] S502, for the (i+1)th iteration, the antigen sequence and antibody sequence i are input into the pre-trained conformation prediction model, and the conformation prediction model performs conformation prediction to obtain the complex conformation i.

[0103] S503, input the complex conformation i into the prediction tools corresponding to various antibody properties respectively, and use the prediction tools of various antibody properties to predict the properties of the complex conformation i to obtain the antibody properties of antibody sequence i.

[0104] S504, obtain the determination criteria for each type of antibody attribute of antibody sequence i.

[0105] S505, compare the attribute value of each type of antibody attribute of antibody sequence i with its respective judgment condition to verify the antibody attribute of antibody sequence i.

[0106] S506, If the antibody properties of antibody sequence i pass the expected verification, antibody sequence i is determined to be the target antibody sequence of the antigen sequence.

[0107] S507, if the antibody properties of antibody sequence i fail the expected verification, generate at least one candidate antibody sequence based on the reference antibody sequence.

[0108] S508, select one or more antibody sequences from at least one candidate antibody sequence as antibody sequence i+1, and continue to execute the i+2th iteration.

[0109] According to the antibody sequence determination method provided in this disclosure, by obtaining a reference antibody sequence for the antigen sequence and iteratively optimizing the reference antibody sequence, antibody R&D costs can be reduced and the success rate of antibody design can be improved. Based on a prediction model, the complex conformation of the antigen and antibody sequences is predicted, and then the antibody properties of the antibody sequence are predicted based on the complex conformation, so as to perform expected verification of antibody properties and achieve simultaneous optimization of multiple antibody properties. When the antibody properties pass the expected verification, the target antibody sequence is obtained; when the antibody properties fail the expected verification, the antibody sequence continues to be iteratively optimized. Through iterative iteration, the efficiency and success rate of antibody design can be improved. By using conformation prediction and property prediction, targeted design optimization of antibodies can be achieved, thereby improving the design success rate and saving a large amount of wet experimental processes, thus reducing design costs.

[0110] like Figure 6 The diagram illustrates the process of optimizing antibody sequences. The optimization process is explained using the (i+1)th iteration as an example. The antigen sequence and antibody sequence i are input into the complex conformation prediction module to predict the complex conformation of antigen and antibody sequences i, resulting in complex conformation i. This complex conformation i is then input into the antibody property prediction module to predict the antibody properties of antibody sequence i. Further, the antibody property prediction module verifies the antibody properties of antibody sequence i. If the verification is successful, antibody sequence i is output as the optimized target antibody sequence; if the verification fails, candidate antibody sequences are generated, and antibody sequence i+1 is selected from these candidate sequences for the (i+2)th iteration.

[0111] For example, to optimize the affinity properties of antibody sequences, a complex conformation prediction model can be used to predict the conformation of the antigen and antibody sequences, thus obtaining the complex conformation. Models such as HelixFold and Alphafold2 can be used to predict the complex conformation. Further, based on the antibody property prediction model, the affinity of the antibody sequence is predicted based on the complex conformation; models such as Foldx can be used to predict affinity. The affinity is then validated according to expectations. After passing the expectation validation, the target antibody sequence is output as the antibody sequence with optimized affinity.

[0112] This disclosure applies to scenarios requiring optimized antibody design based on a given antigen-antibody sequence, including but not limited to the following:

[0113] Antibody affinity maturation: The embodiments disclosed herein can be used to optimize affinity maturation for wild-type antibodies.

[0114] Antibody immunogenicity optimization: The embodiments disclosed herein can be used to optimize the immunogenicity of wild-type antibodies.

[0115] Simultaneous optimization of multiple antibody attributes: Using multiple attribute prediction models and methods, the embodiments disclosed herein can be used to optimize multiple target attributes for wild-type or initial antibodies.

[0116] Corresponding to the antibody sequence determination methods provided in the above embodiments, one embodiment of this disclosure also provides an antibody sequence determination apparatus. Since the antibody sequence determination apparatus provided in this disclosure corresponds to the antibody sequence determination methods provided in the above embodiments, the implementation methods of the above antibody sequence determination methods are also applicable to the antibody sequence determination apparatus provided in this disclosure, and will not be described in detail in the following embodiments.

[0117] Figure 7 This is a schematic diagram of an antibody sequence determination device provided in an embodiment of the present disclosure.

[0118] like Figure 7 As shown, the antibody sequence determination device 700 of this embodiment includes a first iterative optimization module 701, an acquisition module 702, a verification module 703, an output module 704, and a second iterative optimization module 705.

[0119] The first iterative optimization module 701 is used to iteratively optimize the reference antibody sequence, starting from the reference antibody sequence of the antigen sequence.

[0120] The acquisition module 702 is used to acquire, for the (i+1)th iteration, the complex conformation i of the antigen sequence and the antibody sequence i obtained in the i-th iteration, where i is a natural number greater than or equal to 1.

[0121] The verification module 703 is used to obtain the antibody properties of the antibody sequence i based on the complex conformation i, and to perform expected verification on the antibody properties of the antibody sequence i.

[0122] Output module 704 is used to determine that antibody sequence i is the target antibody sequence of the antigen sequence if the antibody properties of antibody sequence i pass the expected verification.

[0123] The second iteration optimization module 705 is used to determine antibody sequence i+1 based on the reference antibody sequence if the antibody properties of antibody sequence i fail the expected verification, and continue to execute the i+2th iteration.

[0124] In one embodiment of this disclosure, the acquisition module 702 is further configured to: input the antigen sequence and the antibody sequence i into a pre-trained conformation prediction model, and have the conformation prediction model perform conformation prediction to obtain the complex conformation i.

[0125] In one embodiment of this disclosure, the verification module 703 is further configured to: input the complex conformation i into prediction tools corresponding to various antibody properties, and have the prediction tools predict the properties of the complex conformation i to obtain the antibody properties of the antibody sequence i.

[0126] In one embodiment of this disclosure, the verification module 703 is further configured to: invoke the antibody property prediction joint model, input the complex conformation i into different prediction network branches in the antibody property prediction joint model, wherein different prediction network branches are applied to predict different antibody properties, and each prediction network branch performs property prediction on the complex conformation i to obtain the antibody properties of the antibody sequence i.

[0127] In one embodiment of this disclosure, the verification module 704 is further configured to: obtain the judgment conditions for each type of antibody attribute of the antibody sequence i; and compare the attribute value of each type of antibody attribute of the antibody sequence i with the respective judgment conditions to verify the antibody attribute of the antibody sequence i.

[0128] In one embodiment of this disclosure, the antibody properties of the antibody sequence i include at least the antibody affinity to the antigen and the antibody immunogenicity.

[0129] In one embodiment of this disclosure, the verification module 703 is further configured to: if the antibody properties of the antibody sequence i fail verification, optimize and fine-tune the antibody property prediction joint model based on the property information of the antibody sequence i and the verification result data.

[0130] In one embodiment of this disclosure, the second iterative optimization module 705 is further configured to: generate at least one candidate antibody sequence based on the reference antibody sequence; and select one or more antibody sequences from the at least one candidate antibody sequence as the antibody sequence i+1.

[0131] In one embodiment of this disclosure, the second iterative optimization module 705 is further configured to: obtain the mutation sites of the reference antibody sequence, and mutate the reference antibody sequence based on the mutation sites to obtain the at least one candidate antibody sequence.

[0132] In one embodiment of this disclosure, the second iterative optimization module 705 is further configured to: obtain the attribute probability distribution of the candidate antibody sequence and the antigen sequence based on the attribute information of the antibody sequence i; and select the antibody sequence i+1 from the at least one candidate antibody sequence based on the attribute probability distribution of the candidate antibody sequence.

[0133] In one embodiment of this disclosure, the second iterative optimization module 705 is further configured to: sample the candidate antibody sequences based on the attribute probability distribution, and obtain one or more antibody sequences ranked first as the antibody sequence i+1.

[0134] According to the antibody sequence determination apparatus provided in this disclosure, by acquiring a reference antibody sequence for an antigen sequence and iteratively optimizing the reference antibody sequence, antibody development costs can be reduced and the success rate of antibody design can be improved. Based on a prediction model, the complex conformation of the antigen and antibody sequences is predicted, and then the antibody properties of the antibody sequence are predicted based on the complex conformation, so as to perform expected verification of antibody properties and achieve simultaneous optimization of multiple antibody properties. When the antibody properties pass the expected verification, the target antibody sequence is obtained; when the antibody properties fail the expected verification, the antibody sequence continues to be iteratively optimized. Through iterative iteration, the efficiency and success rate of antibody design can be improved. By using conformation prediction and property prediction, targeted design optimization of antibodies can be achieved, thereby improving the design success rate and eliminating a large amount of wet experimental processes, saving design costs.

[0135] The acquisition, storage, and application of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0136] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0137] Figure 8 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0138] like Figure 8As shown, device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on computer programs / instructions stored in read-only memory (ROM) 802 or loaded from storage unit 806 into random access memory (RAM) 803. RAM 803 may also store various programs and data required for the operation of device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via bus 804. Input / output (I / O) interface 805 is also connected to bus 804.

[0139] Multiple components in device 800 are connected to I / O interface 805, including: input units 806 such as keyboard, mouse, etc.; output units 807 such as various types of displays, speakers, etc.; storage units 808 such as disks, optical disks, etc.; and communication units 809 such as network cards, modems, wireless transceivers, etc. Communication unit 809 allows device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0140] The computing unit 801 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as the antibody sequence determination method. For example, in some embodiments, the antibody sequence determination method can be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 806. In some embodiments, part or all of the computer program / instructions can be loaded and / or installed on device 800 via ROM 802 and / or communication unit 809. When the computer program / instructions are loaded into RAM 803 and executed by the computing unit 801, one or more steps of the antibody sequence determination method described above can be performed. Alternatively, in other embodiments, the computing unit 801 can be configured to perform the antibody sequence determination method by any other suitable means (e.g., by means of firmware).

[0141] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: implementations in one or more computer programs / instructions that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transferring data and instructions to the storage system, the at least one input device, and the at least one output device.

[0142] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0143] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0144] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0145] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), the Internet, and blockchain networks.

[0146] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. The client-server relationship is created by computer programs / instructions running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0147] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in the disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this document does not impose any restrictions.

[0148] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for determining an antibody sequence, wherein, The method includes: Starting with a reference antibody sequence from an antigen sequence, the reference antibody sequence is iteratively optimized. For the (i+1)th iteration, the antigen sequence and the antibody sequence i are input into the pre-trained conformation prediction model, and the conformation prediction model performs conformation prediction to obtain the complex conformation i, where i is a natural number greater than or equal to 1. Based on the complex conformation i, the antibody properties of the antibody sequence i are obtained, and the antibody properties of the antibody sequence i are verified as expected. If the antibody attribute value of antibody sequence i is greater than the threshold, then through expectation verification, antibody sequence i is determined to be the target antibody sequence of the antigen sequence; If the antibody properties of antibody sequence i fail the expected verification, antibody sequence i+1 is determined based on the reference antibody sequence, and the i+2th iteration is continued.

2. The method according to claim 1, wherein, The step of obtaining the antibody properties of antibody sequence i based on the complex conformation i includes: The complex conformation i is input into the prediction tools corresponding to various antibody properties. The prediction tools predict the properties of the complex conformation i to obtain the antibody properties of the antibody sequence i.

3. The method according to claim 1, wherein, The step of obtaining the antibody properties of antibody sequence i based on the complex conformation i includes: The antibody property prediction joint model is invoked, and the complex conformation i is input into different prediction network branches in the antibody property prediction joint model. Different prediction network branches are applied to predict different antibody properties. Each prediction network branch performs property prediction on the complex conformation i to obtain the antibody property of the antibody sequence i.

4. The method according to claim 1, wherein, The expected verification of the antibody properties of antibody sequence i includes: Obtain the determination criteria for each type of antibody attribute of the antibody sequence i; The attribute value of each type of antibody attribute of antibody sequence i is compared with its respective judgment condition to verify the antibody attribute of antibody sequence i.

5. The method according to claim 4, wherein, The antibody properties of the antibody sequence i include at least the antibody affinity to the antigen and the antibody immunogenicity.

6. The method according to claim 3, wherein, After verifying the antibody properties of antibody sequence i, the process further includes: If the antibody properties of antibody sequence i fail verification, the antibody property prediction joint model is optimized and fine-tuned based on the property information of antibody sequence i and the verification result data.

7. The method according to any one of claims 1-6, wherein, Determining antibody sequence i+1 based on the reference antibody sequence includes: At least one candidate antibody sequence is generated based on the reference antibody sequence; One or more antibody sequences are selected from the at least one candidate antibody sequence as antibody sequence i+1.

8. The method according to claim 7, wherein, The generation of at least one candidate antibody sequence based on the reference antibody sequence includes: The mutation sites of the reference antibody sequence are obtained, and the reference antibody sequence is mutated based on the mutation sites to obtain the at least one candidate antibody sequence.

9. The method according to claim 7, wherein, The step of selecting one or more antibody sequences from the at least one candidate antibody sequence as antibody sequence i+1 includes: Based on the attribute information of the antibody sequence i, obtain the attribute probability distribution of the candidate antibody sequence and the antigen sequence; Based on the attribute probability distribution of the candidate antibody sequences, antibody sequence i+1 is selected from the at least one candidate antibody sequence.

10. The method according to claim 9, wherein, The step of selecting antibody sequence i+1 from the at least one candidate antibody sequence based on the attribute probability distribution of the candidate antibody sequence includes: Based on the attribute probability distribution, the candidate antibody sequences are sampled to obtain one or more antibody sequences that rank highly, which are then used as antibody sequence i+1.

11. An apparatus for determining an antibody sequence, wherein, The device includes: The first iterative optimization module is used to iteratively optimize the reference antibody sequence, starting from the reference antibody sequence of the antigen sequence. The acquisition module is used to input the antigen sequence and the antibody sequence i into a pre-trained conformation prediction model for the (i+1)th iteration, and the conformation prediction model performs conformation prediction to obtain the complex conformation i, where i is a natural number greater than or equal to 1. The verification module is used to obtain the antibody properties of the antibody sequence i based on the complex conformation i, and to perform expected verification of the antibody properties of the antibody sequence i. The output module is used to determine, through expectation verification, that antibody sequence i is the target antibody sequence of the antigen sequence if the antibody attribute value of antibody sequence i is greater than a threshold. The second iteration optimization module is used to determine antibody sequence i+1 based on the reference antibody sequence if the antibody properties of antibody sequence i fail the expected verification, and continue to execute the i+2th iteration.

12. The apparatus according to claim 11, wherein, The verification module is also used for: The complex conformation i is input into the prediction tools corresponding to various antibody properties. The prediction tools predict the properties of the complex conformation i to obtain the antibody properties of the antibody sequence i.

13. The apparatus according to claim 11, wherein, The verification module is also used for: The antibody property prediction joint model is invoked, and the complex conformation i is input into different prediction network branches in the antibody property prediction joint model. Different prediction network branches are applied to predict different antibody properties. Each prediction network branch performs property prediction on the complex conformation i to obtain the antibody property of the antibody sequence i.

14. The apparatus according to claim 11, wherein, The verification module is also used for: Obtain the determination criteria for each type of antibody attribute of the antibody sequence i; The attribute value of each type of antibody attribute of antibody sequence i is compared with its respective judgment condition to verify the antibody attribute of antibody sequence i.

15. The apparatus according to claim 14, wherein, The antibody properties of the antibody sequence i include at least the antibody affinity to the antigen and the antibody immunogenicity.

16. The apparatus according to claim 13, wherein, The verification module is also used for: If the antibody properties of antibody sequence i fail verification, the antibody property prediction joint model is optimized and fine-tuned based on the property information of antibody sequence i and the verification result data.

17. The apparatus according to any one of claims 11-16, wherein, The second iterative optimization module is also used for: At least one candidate antibody sequence is generated based on the reference antibody sequence; One or more antibody sequences are selected from the at least one candidate antibody sequence as antibody sequence i+1.

18. The apparatus according to claim 17, wherein, The second iterative optimization module is also used for: The mutation sites of the reference antibody sequence are obtained, and the reference antibody sequence is mutated based on the mutation sites to obtain the at least one candidate antibody sequence.

19. The apparatus according to claim 17, wherein, The second iterative optimization module is also used for: Based on the attribute information of the antibody sequence i, obtain the attribute probability distribution of the candidate antibody sequence and the antigen sequence; Based on the attribute probability distribution of the candidate antibody sequences, antibody sequence i+1 is selected from the at least one candidate antibody sequence.

20. The apparatus according to claim 19, wherein, The second iterative optimization module is also used for: Based on the attribute probability distribution, the candidate antibody sequences are sampled to obtain one or more antibody sequences that rank highly, which are then used as antibody sequence i+1.

21. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method as described in any one of claims 1-10.

22. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-10.

23. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the method of any one of claims 1-10.

Images