Antigen-antibody interaction potential energy matrix optimization method based on maximizing potential energy difference

Abstract

This disclosure presents an embodiment of a method for optimizing the antigen-antibody interaction potential energy matrix based on maximizing the potential energy difference. One specific implementation of this method includes: determining the average energy value of the natural complex of each training sample in a training sample set based on an initial potential energy matrix; structurally adjusting the crystal structure of the antigen-antibody complex in each training sample in the training sample set to generate a set of non-binding conformations; calculating the energy of each non-binding conformation in the set of non-binding conformations to generate a background average energy value; determining the energy difference between the average energy value of the natural complex and the background average energy value as the potential energy difference; statistically analyzing the residue pair contact frequencies of each non-binding conformation in the set of non-binding conformations to generate a residue frequency matrix; and optimizing the initial potential energy matrix to obtain an optimized potential energy matrix. This implementation can improve the accuracy of predicting antigen-antibody docking posture.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the fields of target optimization technology, antigen-antibody docking technology, and computer technology, and specifically to an antigen-antibody interaction potential energy matrix optimization method based on maximizing potential energy difference. Background Technology

[0002] Antigen-antibody docking is a core technology in immunobiology, drug discovery, and vaccine development. Its goal is to predict the binding patterns (conformations) of antigens and antibodies through computational simulation, providing a basis for understanding immune recognition mechanisms and designing efficient antibody molecules. The core workflow of docking software can be summarized as follows: first, a large number of potential antigen-antibody binding postures (docking conformations) are generated using docking algorithms (such as those based on shape complementarity, fragment assembly, or sampling optimization); then, these postures are scored and ranked using a knowledge-based energy function, ultimately selecting the lowest-energy, most probable physiological binding mode. Knowledge-based potential functions (also known as statistical potential functions) derive their parameters from the statistical regularities of known protein structures, typically focusing on the spatial distribution characteristics (such as distance and angle) of amino acid residue pairs (or atom pairs), and deriving energy parameters through statistical bias. For example, the MJ potential (Miyazawa-Jernigan, Estimation of Effective Interresidue Contact Energies from Protein Crystal Structures: Quasi-Chemical Approximation).

[0003] However, the MJ potential matrix is ​​constructed by statistically analyzing the contact frequency of residue pairs in protein structures and is a general potential function. The antigen-antibody binding has specific characteristics, which makes the MJ potential matrix insufficient in predicting the antigen-antibody docking posture, thus reducing the accuracy of predicting the antigen-antibody docking posture. Summary of the Invention

[0004] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0005] Some embodiments of this disclosure propose an antigen-antibody interaction potential energy matrix optimization method based on maximizing the potential energy difference to solve the technical problems mentioned in the background section above.

[0006] In a first aspect, some embodiments of this disclosure provide a method for optimizing the antigen-antibody interaction potential energy matrix based on maximizing the potential energy difference. The method includes: determining the average energy value of the natural complex of each training sample in a training sample set based on an initial potential energy matrix, wherein the initial potential energy matrix is ​​composed of the initial potential energy difference between each amino acid in a preset amino acid sequence, and the training samples include an antigen-antibody complex crystal structure; structurally adjusting the antigen-antibody complex crystal structure in each training sample in the training sample set to generate a set of non-binding conformations; and calculating the energy of each non-binding conformation in the set of non-binding conformations based on the initial potential energy matrix to generate... The background average energy value; the energy difference between the average energy value of the natural complex and the background average energy value is determined as the potential energy difference; residue pair contact frequency statistics are performed on each unbound conformation in the unbound conformation set to generate a residue frequency matrix; the initial potential energy matrix is ​​optimized based on the potential energy difference, the residue frequency matrix, and hydrophilicity / hydrophobicity parameter constraints to obtain an optimized potential energy matrix, wherein the hydrophilicity / hydrophobicity parameter constraints are used to limit the optimization limits of the energy values ​​in the initial potential energy matrix; in response to the determination that the potential energy difference and the current training round do not meet the preset convergence conditions, the optimized potential energy matrix is ​​used as the initial potential energy matrix, and the initial potential energy matrix is ​​optimized again.

[0007] Secondly, some embodiments of this disclosure provide an antigen-antibody interaction potential energy matrix optimization device based on maximizing potential energy difference. The device includes: a first determining unit configured to determine the average energy value of the natural complex of each training sample in a training sample set based on an initialized potential energy matrix, wherein the initialized potential energy matrix is ​​composed of the initial potential energy difference between each amino acid in a preset amino acid sequence, and the training samples include an antigen-antibody complex crystal structure; a structure adjustment unit configured to adjust the structure of the antigen-antibody complex crystal structure in each training sample in the training sample set to generate a set of non-binding conformations; and an energy calculation unit configured to perform energy calculation on each non-binding conformation in the set of non-binding conformations based on the initialized potential energy matrix to generate a background average energy value. The second determining unit is configured to determine the energy difference between the average energy value of the natural complex and the average energy value of the background as the potential energy difference; the residue pair contact frequency statistics unit is configured to perform residue pair contact frequency statistics on each unbound conformation in the unbound conformation set to generate a residue frequency matrix; the optimization unit is configured to optimize the initial potential energy matrix according to the potential energy difference, the residue frequency matrix and hydrophilicity / hydrophobicity parameter constraints to obtain an optimized potential energy matrix, wherein the hydrophilicity / hydrophobicity parameter constraints are used to limit the optimization limits of the energy values ​​in the initial potential energy matrix; the convergence judgment unit is configured to, in response to determining that the potential energy difference and the current training round do not meet the preset convergence conditions, use the optimized potential energy matrix as the initial potential energy matrix and perform initial potential energy matrix optimization again.

[0008] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0009] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0010] The above embodiments of this disclosure have the following beneficial effects: the antigen-antibody interaction potential energy matrix optimization method based on maximizing potential energy difference in some embodiments of this disclosure can improve the accuracy of predicting antigen-antibody docking posture. Specifically, the reason for the decrease in the accuracy of predicting antigen-antibody docking posture is that the MJ potential energy matrix is ​​constructed by statistically analyzing the contact frequency of residue pairs in the protein structure, and is a general potential energy function. However, antigen-antibody binding has specific characteristics, making the MJ potential energy matrix insufficient for predicting antigen-antibody docking posture. Based on this, the antigen-antibody interaction potential energy matrix optimization method based on maximizing potential energy difference in some embodiments of this disclosure first determines the average energy value of the natural complex of each training sample in the training sample set based on the initial potential energy matrix. The initial potential energy matrix is ​​composed of the initial potential energy difference between each amino acid in a preset amino acid sequence, and the training samples include the crystal structure of the antigen-antibody complex. Here, by determining the average energy value of the natural complex in the training set, it can be used as an optimization benchmark, establishing a benchmark target for subsequent optimization and ensuring that the optimization direction is to reduce the energy of the natural structure. Then, the crystal structures of the antigen-antibody complexes in each training sample in the aforementioned training sample set are structurally adjusted to generate a set of non-binding conformations. Here, generating non-binding conformations provides comparative data for the potential function to learn how to distinguish between correct and incorrect binding modes. Next, based on the aforementioned initialized potential energy matrix, the energy of each non-binding conformation in the aforementioned non-binding conformation set is calculated to generate a background average energy value. Here, quantifying the average energy level of negative samples provides a key reference benchmark for calculating the energy difference. Then, the energy difference between the aforementioned average energy value of the natural complex and the aforementioned background average energy value is determined as the potential energy difference. Here, because the potential energy difference is determined, it can be used as the core optimization objective to optimize the initialized potential energy matrix. Furthermore, residue pair contact frequencies are statistically analyzed for each non-binding conformation in the aforementioned non-binding conformation set to generate a residue frequency matrix. Here, statistically analyzing the frequencies of non-natural interactions from negative samples provides a data-driven basis for updating the initialized potential energy matrix. Then, based on the aforementioned potential energy difference, the aforementioned residue frequency matrix, and the hydrophilicity / hydrophobicity parameter constraints, the initial potential energy matrix is ​​optimized to obtain the optimized potential energy matrix. The hydrophilicity / hydrophobicity parameter constraints are used to limit the optimization limits of the energy values ​​in the initial potential energy matrix. Here, the hydrophilicity / hydrophobicity parameter constraints, introduced based on the chemical properties of amino acids, serve as physical constraints, preventing the optimization process from going astray under purely mathematical drive, and ensuring that the final potential energy parameters conform to basic biophysical laws, thereby greatly improving the model's predictive ability and reliability. Finally, in response to the determination that the aforementioned potential energy difference and the current training round do not meet the preset convergence conditions, the optimized potential energy matrix is ​​used as the initial potential energy matrix, and the initial potential energy matrix is ​​optimized again.Here, the performance (discrimination ability) of the potential energy matrix is ​​continuously converged to the optimal solution through iterative iteration, ensuring the stability and reliability of the final result. This allows the optimized potential energy matrix to possess good specificity when predicting antigen-antibody docking attitudes, thereby improving the accuracy of antigen-antibody docking attitude prediction. Attached Figure Description

[0011] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0012] Figure 1 This is a flowchart of some embodiments of the antigen-antibody interaction potential energy matrix optimization method based on maximizing potential energy difference according to the present disclosure;

[0013] Figure 2 It is the initial potential energy matrix heatmap;

[0014] Figure 3 It is a distance-weighted function curve;

[0015] Figure 4 It is an optimized potential energy matrix heatmap;

[0016] Figure 5 This is a schematic diagram of the structure of some embodiments of the antigen-antibody interaction potential energy matrix optimization device based on maximizing potential energy difference according to the present disclosure;

[0017] Figure 6 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0018] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0019] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0020] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0021] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0022] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0023] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0024] Figure 1 A flowchart 100 is shown, illustrating some embodiments of an antigen-antibody interaction potential energy matrix optimization method based on maximizing potential energy difference according to this disclosure. This antigen-antibody interaction potential energy matrix optimization method based on maximizing potential energy difference includes the following steps:

[0025] Step 101: Based on the initial potential energy matrix, determine the average energy value of the natural complex of each training sample in the training sample set.

[0026] In some embodiments, the execution entity (e.g., a computing device) of the antigen-antibody interaction potential energy matrix optimization method based on maximizing potential energy difference can determine the average energy value of the natural complex of each training sample in the training sample set based on an initialized potential energy matrix. The initialized potential energy matrix is ​​composed of the initial potential energy difference between each amino acid in a preset amino acid sequence, and the training samples include antigen-antibody complex crystal structures.

[0027] As an example, see Figure 2The initial potential energy matrix heatmap shown is a 20×20 initial potential energy matrix heatmap (MJ Matrix Heatmap) composed of the above amino acid sequence (i.e., 20 amino acids). The data of the initial potential energy matrix are the initial potential energy values ​​between each pair of amino acids. Here, the 20 amino acids are: CYS (C) - Cysteine, MET (M) - Methionine, PHE (F) - Phenylalanine, ILE (I) - Isoleucine, LEU (L) - Leucine, VAL (V) - Valine, TRP (W) - Tryptophan, TYR (Y) - Tyrosine, ALA (A) - Alanine, GLY (G) - Glycine, THR (T) - Threonine, SER (S) - Serine, ASN (N) - Asparagine, GLN (Q) - Glutamine, ASP (D) - Aspartic acid, GLU (E) - Glutamic acid, HIS (H) - Histidine, ARG (R) - Arginine, LYS (K) - Lysine, PRO (P) - Proline. Additionally, in the heatmap, smaller values ​​and bluer colors indicate a stronger tendency for the two amino acids to interact, and a more stable interaction; larger values ​​and redder colors indicate a less stable interaction.

[0028] It should be noted that the aforementioned computing devices can be either hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is software, it can be installed on the hardware devices listed above. It can be implemented as, for example, multiple software programs or software modules used to provide distributed services, or as a single software program or software module. No specific limitations are made here.

[0029] In some optional implementations of certain embodiments, the execution entity determines the average energy value of the natural complex of each training sample in the training sample set based on the initialized potential energy matrix, including:

[0030] Step S1: Select amino acid pairs that meet the pairing conditions from the antigen-antibody complex crystal structures of each training sample in the above training sample set to obtain the sample amino acid pair sequence. Specifically, for each antigen-antibody complex crystal structure in the training sample, residue side chain atoms with a distance less than or equal to a preset distance (e.g., 6.5 Å) can be selected as amino acid pairs to obtain the sample amino acid pair set.

[0031] Step S2: Determine the potential energy value corresponding to each amino acid pair in the above sample amino acid pair sequence to obtain the potential energy value sequence. Specifically, the potential energy value (energy value) corresponding to each sample amino acid pair can be selected from the initialized potential energy matrix to obtain the potential energy value sequence.

[0032] Step S3: The average value of each potential energy value in the above potential energy value sequence is determined as the average energy value of the natural complex.

[0033] Step 102: Adjust the crystal structure of the antigen-antibody complex in each training sample in the training sample set to generate a set of non-binding conformations.

[0034] In some embodiments, the execution entity may adjust the crystal structure of the antigen-antibody complex in each training sample in the training sample set to generate a set of non-binding conformations.

[0035] In some optional implementations of certain embodiments, the execution entity performs structural adjustments on the crystal structures of antigen-antibody complexes in each training sample in the training sample set to generate a set of non-binding conformations, including:

[0036] Step S1: For the antigen-antibody complex crystal structure included in each training sample in the above training sample set, perform the following structure adjustment steps:

[0037] The first step is to determine the centroid coordinates of the antigen-antibody complex crystal structures included in the training samples. This can be achieved by determining the geometric center of the atomic coordinates of each α-carbon atom in the crystal structure. The second step involves spatially moving the antigen-antibody complex crystal structures included in the training samples around these centroid coordinates, recording the temporary conformations after each movement to obtain a set of temporary conformations. This can be done by rotating the antigen-antibody complex crystal structures sequentially around the XYZ axes, using the centroid coordinates as the rotation center. The temporary conformation for each rotation (e.g., every 2 degrees) is recorded to obtain the set of temporary conformations.

[0038] The third step involves screening the aforementioned set of temporary conformations to obtain a group of unbound conformations. First, residue side-chain atoms with a distance less than or equal to a preset distance between each temporary conformation are identified as temporary amino acid pairs, resulting in a group of temporary amino acid pairs. Second, temporary conformations corresponding to groups of temporary amino acid pairs with a number greater than or equal to a preset number of temporary amino acid pairs are selected as first-screen conformations, resulting in a group of first-screen conformations. Next, the root mean square deviation (RMSD) of each first-screen conformation in the first-screen conformation group is determined, resulting in a group of root mean square deviations (RMSDs). Then, these RMSD groups are clustered to obtain a set of clusters. Finally, the first-screen conformation with the minimum total potential energy is selected from each cluster in the first-screen conformation set, becoming the set of unbound conformations. Here, the total potential energy is the sum of the potential energy values ​​of the temporary amino acid pairs in the first-screen conformations.

[0039] In practice, screening amino acid pairs can be used to coarsely select conformations with stable interfaces. Secondly, clustering based on root mean square deviation can group first-selection conformations with small structural differences into the same cluster. However, too many first-selection conformations with very small structural differences would be redundant for potential matrix optimization. Therefore, the first-selection conformation with the lowest total potential energy is selected as the representative from each cluster, resulting in a diverse and high-quality set of unbound conformations. This allows for the selection of conformations more suitable for potential matrix optimization, thereby improving the speed and accuracy of potential matrix optimization.

[0040] Step S2 involves determining the obtained groups of unbound conformations as a set of unbound conformations. Specifically, the unbound conformations within each group of unbound conformations can be combined to form a set of unbound conformations.

[0041] Step 103: Based on the initialized potential energy matrix, perform energy calculations on each unbound conformation in the unbound conformation set to generate the background average energy value.

[0042] In some embodiments, the execution entity may perform energy calculations on each unbound conformation in the unbound conformation set based on the initial potential energy matrix to generate a background average energy value.

[0043] In some optional implementations of certain embodiments, the execution entity performs energy calculations on each unbound conformation in the unbound conformation set based on the initial potential energy matrix to generate a background average energy value, including:

[0044] Step S1: Select amino acid pairs that meet the pairing conditions from each unbound conformation in the above unbound conformation set to obtain an amino acid pair sequence set. Specifically, for each unbound conformation, residue side chain atoms with a distance less than or equal to a preset distance can be selected as amino acid pairs to obtain amino acid pair sequences.

[0045] Step S2: Determine the background potential energy value corresponding to each amino acid pair sequence in the above amino acid pair sequence set to obtain the background potential energy value sequence. Specifically, the potential energy value corresponding to each amino acid pair can be selected from the initialized potential energy matrix to obtain the background potential energy value sequence.

[0046] Step S3: Determine the average value of each background potential energy value in the above background potential energy value sequence as the background average energy value.

[0047] Step 104: The energy difference between the average energy value of the natural complex and the average energy value of the background is determined as the potential energy difference.

[0048] In some embodiments, the execution entity may determine the energy difference between the average energy value of the natural complex and the average energy value of the background as a potential energy difference.

[0049] Step 105: Perform residue pair contact frequency statistics on each unbound conformation in the unbound conformation set to generate a residue frequency matrix.

[0050] In some embodiments, the execution entity may perform residue pair contact frequency statistics on each unbound conformation in the unbound conformation set to generate a residue frequency matrix.

[0051] In some optional implementations of certain embodiments, the execution entity performs residue pair contact frequency statistics on each unbound conformation in the aforementioned set of unbound conformations to generate a residue frequency matrix, including:

[0052] Step S1: Determine the Gaussian distance weights for each amino acid in each unbound conformation within the aforementioned unbound conformation set, thus obtaining a Gaussian distance weight set. A Gaussian distance weighting function can be used to enhance the weights of close-range interactions to generate the Gaussian distance weights. For example, for each amino acid, if the distance between the two amino acids in a corresponding amino acid pair is less than or equal to a first preset distance (e.g., 4 Å), increase its contact frequency by one weight (e.g., add 0.95). If the distance between the two amino acids in a pair is greater than the first preset distance and less than or equal to a second preset distance (e.g., 5 Å), increase its contact frequency by one weight (e.g., add 0.5). If the distance between the two amino acids in a pair is greater than the second preset distance and less than or equal to a third preset distance (e.g., 6.5 Å), increase its contact frequency by one weight (e.g., add 0.1). The comprehensive distance weight for each amino acid is then calculated and used as the Gaussian distance weight.

[0053] As an example, such as Figure 3 The graph shows the distance-weighted function. Here, x represents the distance between the two amino acids in an amino acid pair (residue pair distance), and y represents the weight value. Specifically, the peak value (x = 4.0 Å, y ≈ 1.0) in the graph represents the optimal distance for van der Waals interactions, with the lowest energy and strongest interaction. Therefore, it is assigned the highest weight (close to 1). As the distance decreases (x < 4.0 Å): when the atomic distance is less than 4.0 Å, strong van der Waals repulsion occurs, and the interaction becomes unfavorable. Therefore, the weight y decreases from 1. As the distance increases (x ≥ 4.0 Å): as the distance increases, the interaction strength rapidly decreases. The weight y decreases smoothly from 1, gradually approaching 0.

[0054] Step S2: Determine the co-occurrence matrix corresponding to each Gaussian distance weight in the above Gaussian distance weight set. First, arrange the Gaussian distance weights according to the order of the amino acid sequence to form a (20×20) co-occurrence matrix.

[0055] Step S3: Normalize the co-occurrence matrix to obtain the residue frequency matrix.

[0056] Step 106: Based on the potential energy difference, residue frequency matrix and hydrophilicity / hydrophobicity parameter constraints, the initial potential energy matrix is ​​optimized to obtain the optimized potential energy matrix.

[0057] In some embodiments, the execution entity can optimize the initial potential energy matrix based on the potential energy difference, the residue frequency matrix, and the hydrophilicity / hydrophobicity parameter constraint to obtain an optimized potential energy matrix. The hydrophilicity / hydrophobicity parameter constraint is used to limit the optimization bounds of the energy values ​​in the initial potential energy matrix.

[0058] In some optional implementations of certain embodiments, the execution entity optimizes the initial potential energy matrix based on the potential energy difference, the residue frequency matrix, and hydrophilicity / hydrophobicity parameter constraints to obtain an optimized potential energy matrix, including:

[0059] Using the aforementioned residue frequency matrix, potential energy difference, and objective function, the initial potential energy matrix is ​​iteratively processed to obtain the optimized potential energy matrix. The objective function maximizes the potential energy difference, and the initial potential energy matrix is ​​updated using a gradient descent algorithm to obtain the optimized potential energy matrix. Here, the learning rate can be 0.1. The aforementioned hydrophilicity / hydrophobicity parameter constraints can be established based on the amino acid pairs corresponding to each initial potential energy value in the initial potential energy matrix.

[0060] In practice, to prevent excessive deviations from the optimization objective during gradient descent, hydrophilicity / hydrophobicity parameter constraints are introduced. Specifically, for each residue (amino acid) type, parameter constraints can be applied as follows: For hydrophilic residues, a maximum attractive force constraint is applied (e.g., the potential energy value is adjusted upwards (positively) to a range less than or equal to -0.5): This prevents the attractive force between hydrophilic residues (such as ARG, LYS, ASP, GLU, ASN, GLN, SER, THR) from being adjusted too excessively, violating the physical principle that "hydrophilic residues tend to interact with water rather than with each other." For hydrophobic residues, a minimum repulsive force constraint is applied (e.g., the potential energy value is adjusted downwards (negatively) to a range greater than or equal to 0.2): This ensures that the interactions between hydrophobic residues (such as ALA, VAL, ILE, LEU, PHE, TYR, TRP, MET) are always attractive and not incorrectly optimized into repulsion. This ensures that the optimization process is conducted within a reasonable range for antigen-antibody binding, guaranteeing that the final potential energy matrix is ​​not only highly discriminative but also has clear physical meaning.

[0061] As an example, such as Figure 4 The optimized potential energy matrix heatmap shown is shown. Figure 4 In the optimized potential energy matrix (heatmap), the horizontal axis represents the antigen amino acid sequence, and the vertical axis represents the antibody amino acid sequence. Furthermore, the colors in the heatmap characterize the optimized interaction strength between antigen and antibody amino acids. Notably, the interaction strength of the LEU amino acid pair (-19.08) is the strongest, as can be seen from the graph.

[0062] Step 107: In response to the determination that the potential difference and the current training round do not meet the preset convergence conditions, the optimized potential matrix is ​​used as the initial potential matrix, and the initial potential matrix optimization is performed again.

[0063] In some embodiments, the execution entity may, in response to determining that the potential energy difference and the current training round do not meet a preset convergence condition, use the optimized potential energy matrix as the initial potential energy matrix and perform initialization optimization of the potential energy matrix again. The number of iteration rounds is preset (e.g., 50 times). The preset convergence condition may be that the current training round reaches the iteration round, and the number of times the potential energy difference change between adjacent iteration rounds is less than or equal to a preset change threshold is greater than a preset termination number (e.g., 5 times). Here, after every two iterations, the change between two potential energy differences can be determined. If multiple changes are less than or equal to the preset change threshold, iteration convergence is determined. Thus, the initialization optimization of the potential energy matrix can be determined to be complete. The optimized potential energy matrix of the current iteration round is stored. Alternatively, performing initialization optimization of the potential energy matrix again may involve using the optimized potential energy matrix as the initial potential energy matrix and executing steps 101-106 again to obtain a new optimized potential energy matrix. This achieves iterative training of the initial potential energy matrix.

[0064] Furthermore, the optimized potential energy matrix can be tested, showing a 15% increase in the true positive rate and a 20% decrease in the false positive rate. Here, the true positive rate is the probability of a successful prediction. False positive: the probability of a failed prediction. Therefore, the testing steps are as follows: For each natural complex structure in the preset test set, multiple unbound conformations are generated using molecular docking software or a perturbation algorithm. Using the optimized potential energy as a scoring function, the binding interface energy of each test complex and all its corresponding unbound conformations is calculated. Then, if the natural complex conformation ranks in the top N (e.g., the top 10), the prediction is considered successful.

[0065] In practice, the scheme abandons the generalized potential energy function and constructs an iterative optimization framework aimed at maximizing the energy difference. Through gradient descent, the system can automatically learn and adjust the potential energy matrix parameters, significantly enhancing its predictive ability and making it more specific. Then, the scheme cleverly combines data-driven approaches with physical constraints. On the one hand, it learns interaction patterns from large-scale antigen-antibody complex crystal structure data; on the other hand, by introducing physical constraints based on amino acid hydrophilicity and hydrophobicity, it ensures that the optimization process conforms to fundamental principles of biophysics, effectively avoiding overfitting and physically unreasonable results that may arise from purely data-driven approaches. Thus, through the synergistic optimization of machine learning and physical principles, it not only significantly improves the accuracy and discriminative power of the potential energy function in identifying natural conformations but also ensures the physical interpretability of the generated potential energy matrix, providing accurate and reliable predictions for antigen-antibody interactions.

[0066] Optionally, in response to determining that the aforementioned potential energy difference and / or the current training round meets the preset convergence condition, the target antigen-antibody conjugate file is subjected to conformational iteration based on the aforementioned optimized potential energy matrix to generate the optimal binding conformation. Specifically, as shown in the testing steps above, the aforementioned optimized potential energy matrix is ​​used as a scoring function to predict the conformation of the target antigen-antibody conjugate file, thereby obtaining the optimal binding conformation.

[0067] The above embodiments of this disclosure have the following beneficial effects: the antigen-antibody interaction potential energy matrix optimization method based on maximizing potential energy difference in some embodiments of this disclosure can improve the accuracy of predicting antigen-antibody docking posture. Specifically, the reason for the decrease in the accuracy of predicting antigen-antibody docking posture is that the MJ potential energy matrix is ​​constructed by statistically analyzing the contact frequency of residue pairs in the protein structure, and is a general potential energy function. However, antigen-antibody binding has specific characteristics, making the MJ potential energy matrix insufficient for predicting antigen-antibody docking posture. Based on this, the antigen-antibody interaction potential energy matrix optimization method based on maximizing potential energy difference in some embodiments of this disclosure first determines the average energy value of the natural complex of each training sample in the training sample set based on the initial potential energy matrix. The initial potential energy matrix is ​​composed of the initial potential energy difference between each amino acid in a preset amino acid sequence, and the training samples include the crystal structure of the antigen-antibody complex. Here, by determining the average energy value of the natural complex in the training set, it can be used as an optimization benchmark, establishing a benchmark target for subsequent optimization and ensuring that the optimization direction is to reduce the energy of the natural structure. Then, the crystal structures of the antigen-antibody complexes in each training sample in the aforementioned training sample set are structurally adjusted to generate a set of non-binding conformations. Here, generating non-binding conformations provides comparative data for the potential function to learn how to distinguish between correct and incorrect binding modes. Next, based on the aforementioned initialized potential energy matrix, the energy of each non-binding conformation in the aforementioned non-binding conformation set is calculated to generate a background average energy value. Here, quantifying the average energy level of negative samples provides a key reference benchmark for calculating the energy difference. Then, the energy difference between the aforementioned average energy value of the natural complex and the aforementioned background average energy value is determined as the potential energy difference. Here, because the potential energy difference is determined, it can be used as the core optimization objective to optimize the initialized potential energy matrix. Furthermore, residue pair contact frequencies are statistically analyzed for each non-binding conformation in the aforementioned non-binding conformation set to generate a residue frequency matrix. Here, statistically analyzing the frequencies of non-natural interactions from negative samples provides a data-driven basis for updating the initialized potential energy matrix. Then, based on the aforementioned potential energy difference, the aforementioned residue frequency matrix, and the hydrophilicity / hydrophobicity parameter constraints, the initial potential energy matrix is ​​optimized to obtain the optimized potential energy matrix. The hydrophilicity / hydrophobicity parameter constraints are used to limit the optimization limits of the energy values ​​in the initial potential energy matrix. Here, the hydrophilicity / hydrophobicity parameter constraints, introduced based on the chemical properties of amino acids, serve as physical constraints, preventing the optimization process from going astray under purely mathematical drive, and ensuring that the final potential energy parameters conform to basic biophysical laws, thereby greatly improving the model's predictive ability and reliability. Finally, in response to the determination that the aforementioned potential energy difference and the current training round do not meet the preset convergence conditions, the optimized potential energy matrix is ​​used as the initial potential energy matrix, and the initial potential energy matrix is ​​optimized again.Here, the performance (discrimination ability) of the potential energy matrix is ​​continuously converged to the optimal solution through iterative iteration, ensuring the stability and reliability of the final result. This allows the optimized potential energy matrix to possess good specificity when predicting antigen-antibody docking attitudes, thereby improving the accuracy of antigen-antibody docking attitude prediction.

[0068] Further reference Figure 5 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of an antigen-antibody interaction potential energy matrix optimization device based on maximizing the potential energy difference. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, this antigen-antibody interaction potential energy matrix optimization device based on maximizing potential energy difference can be specifically applied to various electronic devices.

[0069] like Figure 5 As shown, an antigen-antibody interaction potential energy matrix optimization device 500 based on maximizing potential energy difference in some embodiments includes: a first determining unit 501, a structure adjustment unit 502, an energy calculation unit 503, a second determining unit 504, a residue pair contact frequency statistics unit 505, an optimization unit 506, and a convergence judgment unit 507. The first determining unit 501 is configured to determine the average energy value of the natural complex of each training sample in the training sample set based on an initial potential energy matrix, wherein the initial potential energy matrix is ​​composed of the initial potential energy difference between each amino acid in a preset amino acid sequence, and the training samples include an antigen-antibody complex crystal structure; the structure adjustment unit 502 is configured to adjust the structure of the antigen-antibody complex crystal structure in each training sample in the training sample set to generate a set of non-binding conformations; the energy calculation unit 503 is configured to perform energy calculation on each non-binding conformation in the set of non-binding conformations based on the initial potential energy matrix to generate a background average energy value; the second determining unit 504 is configured to average the energy of the natural complex... The energy difference between the energy value and the aforementioned background average energy value is determined as the potential energy difference; the residue pair contact frequency statistics unit 505 is configured to perform residue pair contact frequency statistics on each non-binding conformation in the aforementioned non-binding conformation set to generate a residue frequency matrix; the optimization unit 506 is configured to optimize the aforementioned initial potential energy matrix based on the aforementioned potential energy difference, the aforementioned residue frequency matrix, and hydrophilicity / hydrophobicity parameter constraints to obtain an optimized potential energy matrix, wherein the aforementioned hydrophilicity / hydrophobicity parameter constraints are used to limit the optimization limits of the energy values ​​in the initial potential energy matrix; the convergence judgment unit 507 is configured to, in response to determining that the aforementioned potential energy difference and the current training round do not meet the preset convergence conditions, use the aforementioned optimized potential energy matrix as the initial potential energy matrix and perform initial potential energy matrix optimization again.

[0070] It is understandable that the units described in the antigen-antibody interaction potential energy matrix optimization device 500 based on maximizing potential energy difference are related to the reference... Figure 1 The steps described in the method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method also apply to the antigen-antibody interaction potential energy matrix optimization device 500 based on maximizing potential energy difference and the units contained therein, and will not be repeated here.

[0071] The following is for reference. Figure 6 It illustrates a schematic diagram of the structure of an electronic device (such as a computing device) suitable for implementing some embodiments of the present disclosure. Figure 6 The electronic device shown is merely an example and should not be construed as limiting the functionality or scope of the embodiments of this disclosure. Figure 6 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The memory may include a non-volatile storage medium and internal memory. The non-volatile storage medium may store an operating system and a computer program. The computer program includes program instructions that, when executed, cause the processor to perform any of the methods described above. The processor provides computational and control capabilities to support the operation of the entire computer device. The internal memory provides an environment for the execution of the computer program in the non-volatile storage medium; when executed by the processor, the computer program causes the processor to perform any of the methods described above. The network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present disclosure and does not constitute a limitation on the computer device to which the present disclosure is applied. A specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0072] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0073] In one embodiment, the processor is configured to run a computer program stored in a memory to perform the following steps: determining the average energy value of the natural complex of each training sample in the training sample set based on an initial potential energy matrix, wherein the initial potential energy matrix is ​​composed of the initial potential energy difference between each amino acid in a preset amino acid sequence, and the training samples include antigen-antibody complex crystal structures; performing structural adjustments on the antigen-antibody complex crystal structures in each training sample in the training sample set to generate a set of unbound conformations; and performing energy calculations on each unbound conformation in the set of unbound conformations based on the initial potential energy matrix to generate a background average. Energy value; the energy difference between the average energy value of the above-mentioned natural complex and the average energy value of the above-mentioned background is determined as the potential energy difference; residue pair contact frequency statistics are performed on each unbound conformation in the above-mentioned unbound conformation set to generate a residue frequency matrix; the above-mentioned initial potential energy matrix is ​​optimized according to the above-mentioned potential energy difference, the above-mentioned residue frequency matrix and hydrophilicity / hydrophobicity parameter constraints to obtain an optimized potential energy matrix, wherein the above-mentioned hydrophilicity / hydrophobicity parameter constraints are used to limit the optimization limit of the energy value in the initial potential energy matrix; in response to determining that the above-mentioned potential energy difference and the current training round do not meet the preset convergence condition, the above-mentioned optimized potential energy matrix is ​​used as the initial potential energy matrix, and the initial potential energy matrix is ​​optimized again.

[0074] This disclosure also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, and the method implemented when the program instructions are executed can be referred to the various embodiments of the methods described above.

[0075] The aforementioned computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. Alternatively, the aforementioned computer-readable storage medium may be an external storage device of the computer device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.

[0076] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0077] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for optimizing the potential energy matrix of antigen-antibody interaction based on the maximization of potential energy difference, characterized in that, The method comprises the following steps: determining the natural complex average energy value of each training sample in the training sample set based on the initial potential energy matrix, wherein the initial potential energy matrix is composed of the initial potential energy difference between each amino acid in the preset amino acid sequence, and the training sample comprises an antigen-antibody complex crystal structure; adjusting the structure of the antigen-antibody complex crystal structure in each training sample in the training sample set to generate a non-binding conformation set; based on the initial potential energy matrix, energy calculation is performed on each non-binding conformation in the non-binding conformation set to generate a background average energy value; determining the energy difference between the natural complex average energy value and the background average energy value as a potential energy difference; statistical analysis of the residue docking contact frequency of each non-binding conformation in the non-binding conformation set to generate a residue frequency matrix; According to the potential energy difference, the residue frequency matrix and the hydrophilic-hydrophobic parameter constraint, the initial potential energy matrix is optimized to obtain an optimized potential energy matrix, wherein the hydrophilic-hydrophobic parameter constraint is used to limit the optimization limit of the energy value in the initial potential energy matrix; In response to determining that the potential energy difference and the current training round do not satisfy the preset convergence condition, the optimized potential energy matrix is used as the initial potential energy matrix, and the initial potential energy matrix optimization is performed again; In response to determining that the potential energy difference and / or the current training round satisfies the preset convergence condition, performing conformation prediction on a target antigen-antibody binder file based on the optimized potential energy matrix to generate an optimal binding conformation.

2. The method of claim 1, wherein, The method comprises the following steps: selecting an amino acid pair that satisfies a pairing condition from the antigen-antibody complex crystal structure in each training sample in the training sample set to obtain a sample amino acid pair group sequence; determining the potential energy value corresponding to each sample amino acid pair group in the sample amino acid pair group sequence to obtain a potential energy value sequence; determining the average value of each potential energy value in the potential energy value sequence as the natural complex average energy value.

3. The method of claim 2, wherein, The method comprises the following steps: For the antigen-antibody complex crystal structure included in each training sample in the training sample set, the following structure adjustment steps are performed: determining the antigen-antibody centroid coordinates in the antigen-antibody complex crystal structure included in the training sample; spatially moving the antigen-antibody complex crystal structure included in the training sample with the antigen-antibody centroid coordinates as the center to record the temporary conformations after each movement to obtain a temporary conformation set; screening the temporary conformation set to obtain a non-binding conformation group; determining each obtained non-binding conformation group as a non-binding conformation set.

4. The method of claim 3, wherein, The method comprises the following steps: screening the non-binding conformation set to obtain an amino acid pair sequence set; determine a background potential energy value corresponding to each amino acid pair sequence in the amino acid pair sequence set, to obtain a background potential energy value sequence; determine an average value of each background potential energy value in the background potential energy value sequence as a background average energy value.

5. The method of claim 4, wherein, The residue contact frequency statistics unit is configured to perform residue pair contact frequency statistics on each of the non-binding conformations in the non-binding conformation set to generate a residue frequency matrix. determine a Gaussian distance weight of each amino acid in each of the non-binding conformations in the non-binding conformation set to obtain a Gaussian distance weight set; determine a co-occurrence matrix corresponding to each Gaussian distance weight in the Gaussian distance weight set; normalize the co-occurrence matrix to obtain a residue frequency matrix.

6. The method of claim 5, wherein, The optimization unit is configured to optimize the initialization potential matrix according to the potential difference, the residue frequency matrix, and a hydrophilic-hydrophobic parameter constraint to obtain an optimized potential matrix, wherein the hydrophilic-hydrophobic parameter constraint is used to limit the optimization range of the energy value in the initialization potential matrix. The optimization unit is configured to optimize the initialization potential matrix according to the potential difference, the residue frequency matrix, and a hydrophilic-hydrophobic parameter constraint to obtain an optimized potential matrix, wherein the hydrophilic-hydrophobic parameter constraint is used to limit the optimization range of the energy value in the initialization potential matrix.

7. An apparatus for optimizing a potential energy matrix of an antigen-antibody interaction based on maximization of potential energy difference, characterized in that, The optimization unit is configured to optimize the initialization potential matrix according to the potential difference, the residue frequency matrix, and a hydrophilic-hydrophobic parameter constraint to obtain an optimized potential matrix, wherein the hydrophilic-hydrophobic parameter constraint is used to limit the optimization range of the energy value in the initialization potential matrix. The optimization unit is configured to optimize the initialization potential matrix according to the potential difference, the residue frequency matrix, and a hydrophilic-hydrophobic parameter constraint to obtain an optimized potential matrix, wherein the hydrophilic-hydrophobic parameter constraint is used to limit the optimization range of the energy value in the initialization potential matrix. The convergence judgment unit is configured to, in response to determining that the potential difference and the current training round do not satisfy a preset convergence condition, take the optimized potential matrix as the initialization potential matrix and perform initialization potential matrix optimization again. In response to determining that the potential difference and / or the current training round satisfy the preset convergence condition, perform conformation prediction on a target antigen-antibody binder file based on the optimized potential matrix to generate an optimal binding conformation. one or more processors; a storage device having one or more programs stored thereon, ​ ​ ​ 8. An electronic device, comprising: ​ ​ ​ When the one or more programs are executed by the one or more processors, the one or more processors implement a method as claimed in any one of claims 1-6.

9. A computer readable medium characterized by A computer program product having stored thereon a computer program, wherein the program, when executed by a processor, implements a method as claimed in any one of claims 1-6.

Images