AAV capsid protein design method, system, device and storage medium

Abstract

The present invention discloses a design method, system, device and storage medium for AAV capsid protein, including: setting an initial temperature according to a preset temperature, the initial temperature characterizing the mutation probability of each position of the initial AAV sequence; determining a cooling rate according to the sum of the improvement rate of the targeting score and the yield score of the current AAV capsid protein sequence, the diversity score and a preset maximum cooling rate; determining the current temperature according to the initial temperature and the cooling rate, selecting a first preset number of AAV capsid protein sequences at the current temperature according to an objective function, and continuing cooling optimization until the preset index requirements are met, and stopping the cooling search; selecting a second preset number of AAV capsid protein sequences from all AAV capsid protein sequences that meet the preset index requirements for clustering to determine the final designed AAV capsid protein sequence. The embodiments of the present invention can reduce the risk of being trapped in a local optimal solution, improve the targeting and yield of AAV capsid proteins, and can be widely used in the field of bioinformatics.

Description

Technical Field

[0001] The present invention relates to the field of bioinformatics, and in particular to a design method, system, device and storage medium for AAV capsid protein. Background Art

[0002] Adeno-associated virus (AAV) is a single-stranded, linear, defective virus. Due to its low genotoxicity in humans, it is one of the most promising gene therapy vectors. However, wild-type AAV suffers from limited tissue-specific targeting and production, necessitating engineering of the AAV capsid protein to achieve optimal efficacy.

[0003] There are many approaches to designing AAV capsid proteins, including natural discovery, rational design, directed evolution, and computational design. Computational design is an emerging approach that uses artificial intelligence and machine learning technologies to perform diverse design and sequence variation on AAV capsid proteins, thereby guiding the design of both functionality and diversity.

[0004] Heuristic algorithms are common computational design methods. They generate a feasible solution for each instance of a combinatorial optimization problem within acceptable computational time and space. The degree to which this feasible solution deviates from the optimal solution is generally unpredictable. A drawback of heuristic algorithms is that they cannot guarantee that the resulting feasible solution is identical to the optimal solution. In most cases, it is even impossible to determine the degree to which the resulting solution approximates the optimal solution. Furthermore, because they typically seek to solve the problem from a local perspective, they can easily become trapped in a local optimum.

[0005] The design method based on machine learning models is also a type of computational design method. It can quickly predict the structure and stability of AAV capsid proteins without requiring a large amount of computing resources and time. It predicts the structure and stability of new AAV capsid proteins by training on known AAV capsid protein sequences and structures. There are many types of machine learning models, such as neural networks, random forests, gradient boosting trees, and so on. Since it is difficult for training data to cover the entire possibility space, the design method based on machine learning models is prone to falling into local optimal solutions. Therefore, when using computational design methods to design AAV capsid proteins, it is easy to fall into local optimal solutions during the design process, which often results in the designed AAV amino acid sequences having a very high similarity, or even amino acid sequences with synonymous mutations, which will cause all the designed AAV capsid proteins to fail to achieve the desired effect (such as strong tissue-specific targeting and high yield). Summary of the Invention

[0006] In view of this, the purpose of the embodiments of the present invention is to provide a design method, system, device and storage medium for AAV capsid protein, which can reduce the risk of being trapped in local optimal solutions and improve the targeting and yield of AAV capsid protein.

[0007] In a first aspect, an embodiment of the present invention provides a method for designing an AAV capsid protein, comprising the following steps:

[0008] The AAV capsid protein sequence to be optimized is randomly initialized, and the initial temperature is set according to a preset temperature. The initial temperature represents the mutation possibility of each position of the initial AAV sequence;

[0009] The cooling rate is determined based on the sum of the improvement rate of the targeting score and the yield score of the current AAV capsid protein sequence compared with the wild-type AAV capsid protein sequence, the diversity score of the current AAV capsid protein sequence, and the preset maximum cooling rate;

[0010] Determining the current temperature based on the previous temperature and the cooling rate, selecting a first preset number of AAV capsid protein sequences at the current temperature according to an objective function and continuing the cooling optimization until the AAV capsid protein sequences at the current temperature meet the preset index requirements, and then stopping the cooling search; the objective function includes a targeting score, a yield score, and a diversity score;

[0011] A second preset number of undetermined AAV capsid protein sequences are selected from all AAV capsid protein sequences that meet the preset index requirements, the undetermined AAV capsid protein sequences are clustered, and a third preset number of AAV capsid protein sequences are selected from each type of sequence as the final designed AAV capsid protein sequences.

[0012] Optionally, the diversity score of the AAV capsid protein sequence is obtained by the following steps:

[0013] Obtaining a batch of AAV capsid protein sequences, and clustering the AAV capsid protein sequences according to the similarity and coverage between the AAV capsid protein sequences;

[0014] The diversity score of each class of AAV capsid protein sequences was determined based on the total number of classes in the class.

[0015] Optionally, clustering the AAV capsid protein sequences according to the similarity and coverage between the AAV capsid protein sequences includes:

[0016] Determine query and target sequences from a batch of AAV capsid protein sequences;

[0017] The similarity is determined based on the ratio of the number of identical amino acids to the total number of amino acids in the query and target sequences;

[0018] The coverage is determined based on the ratio of the length of the longest common subsequence between the query sequence and the target sequence to the total length of the query sequence;

[0019] Sequences that meet both similarity and coverage requirements are clustered into the same category.

[0020] Alternatively, the diversity score of the AAV capsid protein sequence is calculated using the following formula:

[0021]

[0022] Alternatively, the targeting score and yield score of the AAV capsid protein sequence are obtained by the following steps:

[0023] The AAV capsid protein sequence was input into the trained estimation model to obtain the targeting score and yield score.

[0024] Optionally, the cooling rate is determined based on the sum of the improvement rate of the targeting score and the yield score of the current AAV capsid protein sequence compared to the wild-type AAV capsid protein sequence, the diversity score of the current AAV capsid protein sequence, and a preset maximum cooling rate, specifically including:

[0025] determining a first improvement rate based on a targeting score of the current AAV capsid protein sequence and a targeting score of a wild-type AAV capsid protein sequence;

[0026] determining a second improvement rate based on the yield score of the current AAV capsid protein sequence and the yield score of the wild-type AAV capsid protein sequence;

[0027] Determine an average value based on the first improvement rate and the second improvement rate, and calculate the ratio of the difference between 1 and the average value to the diversity score of the current AAV capsid protein sequence;

[0028] The smaller value between the ratio and the preset maximum cooling rate is taken as the cooling rate.

[0029] Optionally, selecting a first preset number of AAV capsid protein sequences at the current temperature according to the objective function to continue the cooling optimization specifically includes:

[0030] Mutate each position of the AAV capsid protein sequence according to the current temperature;

[0031] Calculate the objective function value of all mutated AAV capsid protein sequences at the current temperature;

[0032] Recalculating the cooling rate of the AAV capsid protein sequence and updating the current temperature of the AAV capsid protein sequence;

[0033] The function values ​​are sorted by high to low, and the AAV capsid protein sequences with the first preset number of function value rankings are selected to continue the cooling optimization.

[0034] In a second aspect, an embodiment of the present invention provides a system for designing an AAV capsid protein, comprising:

[0035] The first module is used to randomly initialize the AAV capsid protein sequence to be optimized and set the initial temperature according to a preset temperature. The initial temperature represents the mutation possibility of each position of the initial AAV sequence;

[0036] The second module is used to determine the cooling rate according to the sum of the improvement rate of the targeting score and the yield score of the current AAV capsid protein sequence compared with the wild-type AAV capsid protein sequence, the diversity score of the current AAV capsid protein sequence and the preset maximum cooling rate;

[0037] The third module is used to determine the current temperature based on the temperature at the previous moment and the cooling rate, and select a first preset number of AAV capsid protein sequences at the current temperature according to the objective function to continue cooling optimization until the AAV capsid protein sequences at the current temperature meet the preset index requirements, and then stop the cooling search; the objective function includes the targeting score, the yield score and the diversity score;

[0038] The fourth module is used to select a second preset number of undetermined AAV capsid protein sequences from all AAV capsid protein sequences that meet the preset index requirements, cluster the undetermined AAV capsid protein sequences, and select a third preset number of AAV capsid protein sequences from each type of sequence as the final designed AAV capsid protein sequences.

[0039] In a third aspect, an embodiment of the present invention provides a design device for an AAV capsid protein, comprising:

[0040] at least one processor;

[0041] at least one memory for storing at least one program;

[0042] When the at least one program is executed by the at least one processor, the at least one processor implements the above method.

[0043] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium storing a program executable by a processor, wherein the program executable by the processor is used to perform the above method when executed by the processor.

[0044] The implementation of the embodiment of the present invention includes the following beneficial effects: This embodiment adopts a simulated annealing algorithm to design AAV capsid protein, determines the objective function according to the targeting score, yield score and diversity score, and constrains the designed AAV capsid protein sequence according to the objective function, which can avoid falling into a local optimal solution and causing the designed AAV capsid protein amino acid sequence to be too similar, reducing the risk of failure in the verification performance and actual application of the designed AAV capsid protein, and screening and diversity clustering analysis of all AAV capsid protein sequences that meet the preset index requirements, again avoiding the AAV capsid protein sequences finally screened out to be too similar, avoiding verification risks, thereby improving the targeting and yield of the designed AAV capsid protein and achieving ideal results. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 1 is a schematic flow chart of the steps of a method for designing an AAV capsid protein provided by an embodiment of the present invention;

[0046] Figure 2 This is a schematic flow chart of the steps of a temperature updating method provided by an embodiment of the present invention;

[0047] Figure 3 1 is a schematic flow chart of steps for determining the diversity score of an AAV capsid protein sequence provided by an embodiment of the present invention;

[0048] Figure 4 1 is a schematic flow chart of the steps of a clustering method for AAV capsid protein sequences provided in an embodiment of the present invention;

[0049] Figure 5 This is a schematic flow chart of steps for determining a cooling rate according to an embodiment of the present invention;

[0050] Figure 6 This is a schematic flow chart of the steps for selecting an AAV capsid protein sequence for temperature reduction optimization provided by an embodiment of the present invention;

[0051] Figure 7 This is a structural block diagram of a design system for an AAV capsid protein provided by an embodiment of the present invention;

[0052] Figure 8 This is a structural block diagram of a design device for AAV capsid protein provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The step numbers in the following embodiments are provided for ease of description only and do not limit the order of the steps. The order of execution of the steps in the embodiments can be adaptively adjusted based on the understanding of those skilled in the art.

[0054] The simulated annealing algorithm is a stochastic optimization algorithm designed to mimic the physical annealing process. It uses a similar process to physical annealing, starting at a high temperature (equivalent to the algorithm's random search), followed by gradual annealing. At each temperature (equivalent to each state transition in the algorithm), the algorithm gradually cools down (equivalent to the algorithm's local search), ultimately reaching the physical ground state (equivalent to the algorithm finding the optimal solution).

[0055] High temperature process: Enhance the thermal motion of particles and make them deviate from the equilibrium position, with the aim of eliminating any non-uniform state that may have existed in the system.

[0056] Isothermal process: During the annealing process, the temperature is gradually reduced to reach a thermal equilibrium state at each temperature. For a closed system that exchanges heat with the surrounding environment while maintaining a constant temperature, the law of less free energy is satisfied. The spontaneous change of the system state always proceeds in the direction of reducing free energy. When the free energy reaches a minimum, the system reaches equilibrium.

[0057] Cooling process: The thermal motion of particles is weakened and gradually ordered, and the energy of the system gradually decreases, thus obtaining a low-energy crystal structure. The annealing process is completed when the liquid solidifies into a solid crystalline state.

[0058] like Figure 1 As shown, an embodiment of the present invention provides a method for designing an AAV capsid protein, which includes the following steps.

[0059] S100, randomly initializing the AAV capsid protein sequence to be optimized, setting the initial temperature according to a preset temperature, and the initial temperature represents the mutation possibility of each position of the initial AAV sequence.

[0060] The initial temperature is the initial high temperature during the high temperature process of the simulated annealing algorithm. The higher the initial temperature, the greater the possibility of mutation at each position of the AAV capsid protein sequence. Specifically, the AAV capsid protein sequence to be optimized is first randomly initialized, and then the initial temperature is determined based on the possibility of mutation at each position of the AAV capsid protein sequence and the actual application requirements. It should be noted that the number of AAV capsid protein sequences to be optimized is determined according to the actual application, and this embodiment does not impose any specific restrictions.

[0061] S200, determining a cooling rate based on a preset maximum cooling rate of the diversity score of the current AAV capsid protein sequence, the sum of the improvement rates of the targeting score and the yield score of the current AAV capsid protein sequence compared to the wild-type AAV capsid protein sequence, and the current AAV capsid protein sequence.

[0062] The initial mutation probability indicates the probability that an amino acid in the AAV capsid protein sequence mutates into any amino acid. It should be noted that the initial mutation probability is determined based on actual application and is not specifically limited in this embodiment. For example, the initial mutation probability is set to 0.7.

[0063] It should be noted that the preset maximum cooling rate is determined according to actual application and is not specifically limited in this embodiment. For example, the preset maximum cooling rate is set to 0.7.

[0064] The current AAV capsid protein sequence refers to the AAV capsid protein sequence at the current temperature. The real-time temperature gradually cools down from the initial temperature. During the cooling process, the AAV capsid protein sequence will mutate. Different cooling rates correspond to different mutated AAV capsid protein sequences. In this embodiment, a suitable cooling rate is determined by comprehensively considering the sum of the targeting score and the improvement rate of the yield score of the current AAV capsid protein sequence, the diversity score, and the preset maximum cooling rate to find an AAV capsid protein sequence that meets the requirements.

[0065] S300. Determine the current temperature based on the temperature at the previous moment and the cooling rate, select the first preset number of AAV capsid protein sequences at the current temperature according to the objective function to continue cooling optimization until the AAV capsid protein sequence at the current temperature meets the preset index requirements, and stop the cooling search; the objective function includes the targeting score, the yield score and the diversity score.

[0066] It should be noted that the current temperature = the temperature at the previous moment * the cooling rate. The objective function includes multiple factors such as the targeting score, the yield score, and the diversity score. The weight of each factor is determined according to the actual application, and this embodiment does not impose any specific restrictions. For example: objective function = targeting score + yield score + diversity score. The first preset number is determined according to the actual application, and this embodiment does not impose any specific restrictions. The preset indicator requirements are determined according to the actual application, and this embodiment does not impose any specific restrictions. For example, the preset indicator requirement is set to the sum of the improvement rates of the targeting score and the yield score reaching 200% to achieve the set high-target and high-yield requirements.

[0067] Specifically, see Figure 2 , determine the current temperature according to the temperature at the previous moment and the cooling rate, calculate the index data of the AAV capsid protein sequence at the current temperature, and stop the cooling search if the index data meets the preset index requirements; if the index data does not meet the preset index requirements, select the first preset number of AAV capsid protein sequences with potentially better performance according to the objective function to continue cooling optimization until the index data meets the preset index requirements and stop the cooling search.

[0068] S400. Select a second preset number of undetermined AAV capsid protein sequences from all AAV capsid protein sequences that meet preset index requirements, cluster the undetermined AAV capsid protein sequences, and select a third preset number of AAV capsid protein sequences from each type of sequence as the final designed AAV capsid protein sequences.

[0069] It should be noted that the second preset number and the third preset number are determined according to actual applications and are not specifically limited in this embodiment. For example, the second preset number is set to 500 and the third preset number is set to 5.

[0070] Specifically, first, a second preset number of pending AAV capsid protein sequences are selected from all AAV capsid protein sequences that meet the preset index requirements accumulated at multiple temperatures, and then the selected pending AAV capsid protein sequences are clustered. Then, a third preset number of AAV capsid protein sequences with better performance indicators are selected from each type of sequence as target AAV capsid protein sequences.

[0071] Optionally, see Figure 3 The diversity score of AAV capsid protein sequences is obtained by the following steps:

[0072] S210, obtaining a batch of AAV capsid protein sequences, and clustering the AAV capsid protein sequences according to the similarity and coverage between the AAV capsid protein sequences;

[0073] S220. Determine a diversity score of each class of AAV capsid protein sequences according to the total number of classes in the class of AAV capsid protein sequences.

[0074] Specifically, the AAV capsid protein sequences were first clustered according to the similarity and coverage between the AAV capsid protein sequences, and the sequences with higher similarity and coverage were clustered into one category; then the diversity score of each category of AAV capsid protein sequences was calculated based on the total number of categories in the sequence. The greater the total number of categories in each category of AAV capsid protein sequences, the lower the diversity score.

[0075] Optionally, see Figure 4 , clustering of AAV capsid protein sequences based on their similarity and coverage, specifically including:

[0076] S211, determining a query sequence and a target sequence from a batch of AAV capsid protein sequences;

[0077] S212, determining similarity based on the ratio of the number of identical amino acids to the total number of amino acids in the query sequence and the target sequence;

[0078] S213, determining coverage based on a ratio of the length of the longest common subsequence between the query sequence and the target sequence to the total length of the query sequence;

[0079] S214. Grouping sequences that meet both similarity requirements and coverage requirements into the same category.

[0080] Specifically, first determine the target sequence in the batch of AAV capsid protein sequences, and use other sequences as query sequences; then find the number of identical amino acids between the target sequence and the query sequence, and determine the similarity by the ratio of the number of identical amino acids in the query sequence and the target sequence to the total number of amino acids; then find the longest common subsequence length between the query sequence and the target sequence, and determine the coverage by the ratio of the longest common subsequence length between the query sequence and the target sequence to the total length of the query sequence; finally, cluster the sequences that exceed both the similarity requirement and the coverage requirement into the same category. It should be noted that the similarity requirement and the coverage requirement can be determined according to actual application, and this embodiment does not impose specific restrictions. For example, sequences with similarity and coverage exceeding 90% are clustered into the same category.

[0081] Alternatively, the diversity score of the AAV capsid protein sequence is calculated using the following formula:

[0082]

[0083] It should be noted that the total number of classes represents the total number of AAV capsid protein sequence classes within each class. The greater the total number of AAV capsid protein sequence classes within a sequence of the same type, the lower the diversity score; the fewer the total number of AAV capsid protein sequence classes within a sequence of the same type, the higher the diversity score.

[0084] Alternatively, the targeting score and yield score of the AAV capsid protein sequence are obtained by the following steps:

[0085] The AAV capsid protein sequence was input into the trained estimation model to obtain the targeting score and yield score.

[0086] It should be noted that before using the prediction model, it is necessary to train it with AAV capsid protein sequence sample data until the training meets the requirements and obtain the prediction model. The AAV capsid protein sequence sample data includes AAV capsid protein sequence samples and their corresponding targeting scores and yield scores.

[0087] Optionally, see Figure 5 The cooling rate is determined based on the sum of the improvement rate of the current AAV capsid protein sequence's targeting score and yield score compared to the wild-type AAV capsid protein sequence, the current AAV capsid protein sequence's diversity score, and the initial mutation probability, including:

[0088] S230, determining a first improvement rate according to the targeting score of the current AAV capsid protein sequence and the targeting score of the wild-type AAV capsid protein sequence;

[0089] S240, determining a second improvement rate based on the yield score of the current AAV capsid protein sequence and the yield score of the wild-type AAV capsid protein sequence;

[0090] S250, determining an average value according to the first improvement rate and the second improvement rate, and calculating a ratio of a difference between 1 and the average value to a diversity score of the current AAV capsid protein sequence;

[0091] S260: Take the smaller value of the ratio and the preset maximum cooling rate as the cooling rate.

[0092] Specifically, first, the targeting score and yield score of the current AAV capsid protein sequence and the wild-type AAV capsid protein sequence are calculated by the estimation model respectively; then, the first improvement rate is determined based on the targeting score of the current AAV capsid protein sequence and the targeting score of the wild-type AAV capsid protein sequence, and the second improvement rate is determined based on the yield score of the current AAV capsid protein sequence and the yield score of the wild-type AAV capsid protein sequence; then, the cooling rate is determined based on the first improvement rate, the second improvement rate, the diversity score and the preset maximum cooling rate. The calculation formulas for the first improvement rate and the second improvement rate are as follows:

[0093] First improvement rate = (targeting score of current AAV capsid protein sequence - targeting score of wild-type AAV capsid protein sequence) / targeting score of wild-type AAV capsid protein sequence

[0094] Second improvement rate = (yield score of current AAV capsid protein sequence - yield score of wild-type AAV capsid protein sequence) / yield score of wild-type AAV capsid protein sequence

[0095] The cooling rate is calculated as follows: cooling rate = min([1-(sum of improvement rates / 2)] / diversity score, with a preset maximum cooling rate).

[0096] Optionally, see Figure 6 , according to the objective function, the first preset number of AAV capsid protein sequences at the current temperature are selected to continue the cooling optimization, specifically including:

[0097] S310, mutating each position of the AAV capsid protein sequence according to the current temperature;

[0098] S320, calculating the objective function values ​​of all mutated AAV capsid protein sequences at the current temperature;

[0099] S330, recalculating the cooling rate of the AAV capsid protein sequence and updating the current temperature of the AAV capsid protein sequence;

[0100] S340, sorting the function values ​​by high and low, and selecting the AAV capsid protein sequences with the first preset number of function value rankings to continue cooling optimization.

[0101] Specifically, the function values ​​corresponding to all AAV capsid protein sequences at the current temperature are first calculated according to the calculation formula of the objective function, and then the function values ​​are sorted by high to low order. The AAV capsid protein sequences with the first preset number of function value rankings are selected to continue the cooling optimization, and the remaining AAV capsid protein sequences ranked at the bottom are discarded.

[0102] The implementation of the embodiment of the present invention includes the following beneficial effects: This embodiment adopts a simulated annealing algorithm to design AAV capsid protein, determines the objective function according to the targeting score, yield score and diversity score, and constrains the designed AAV capsid protein sequence according to the objective function, which can avoid falling into a local optimal solution and causing the designed AAV capsid protein amino acid sequence to be too similar, reducing the risk of failure in the verification performance and actual application of the designed AAV capsid protein, and screening and diversity clustering analysis of all AAV capsid protein sequences that meet the preset index requirements, again avoiding the AAV capsid protein sequences finally screened out to be too similar, avoiding verification risks, thereby improving the targeting and yield of the designed AAV capsid protein and achieving ideal results.

[0103] like Figure 7 As shown, the embodiment of the present invention provides a design system for AAV capsid protein, comprising:

[0104] The first module is used to randomly initialize the AAV capsid protein sequence to be optimized, and set the initial temperature according to the preset initial temperature, which is the mutation probability of each position of the initial AAV sequence;

[0105] The second module is used to determine the cooling rate according to the sum of the improvement rate of the targeting score and the yield score of the current AAV capsid protein sequence compared with the wild-type AAV capsid protein sequence, the diversity score of the current AAV capsid protein sequence and the preset maximum cooling rate;

[0106] The third module is used to determine the current temperature based on the temperature at the previous moment and the cooling rate, and select a first preset number of AAV capsid protein sequences at the current temperature according to the objective function to continue cooling optimization until the AAV capsid protein sequences at the current temperature meet the preset index requirements, and then stop the cooling search; the objective function includes the targeting score, the yield score and the diversity score;

[0107] The fourth module is used to select a second preset number of undetermined AAV capsid protein sequences from all AAV capsid protein sequences that meet the preset index requirements, cluster the undetermined AAV capsid protein sequences, and select a third preset number of AAV capsid protein sequences from each type of sequence as the final designed AAV capsid protein sequences.

[0108] It can be seen that the contents of the above method embodiments are all applicable to the present system embodiments. The functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0109] like Figure 8 As shown, an embodiment of the present invention provides a design device for AAV capsid protein, comprising:

[0110] at least one processor;

[0111] at least one memory for storing at least one program;

[0112] When the at least one program is executed by the at least one processor, the at least one processor implements the above method.

[0113] Among them, the memory is a non-transient computer-readable storage medium that can be used to store non-transient software programs and non-transient computer executable programs. The memory may include a high-speed random access memory and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory optionally includes a remote memory remotely arranged relative to the processor, and these remote memories can be connected to the processor via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network and a combination thereof.

[0114] It can be seen that the contents of the above method embodiments are all applicable to the present device embodiments. The functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0115] In addition, the embodiments of the present application further disclose a computer program product or computer program, which is stored in a computer-readable storage medium. The processor of a computer device can read the computer program from the computer-readable storage medium, and the processor executes the computer program, so that the computer device performs the above-mentioned method. Similarly, the contents of the above-mentioned method embodiment are all applicable to the present storage medium embodiment, and the functions specifically implemented by the present storage medium embodiment are the same as those of the above-mentioned method embodiment, and the beneficial effects achieved are also the same as those achieved by the above-mentioned method embodiment.

[0116] An embodiment of the present invention further provides a computer-readable storage medium, which stores a program executable by a processor. The program executable by the processor is used to implement the above method when executed by the processor.

[0117] It is understood that all or some steps, systems in the disclosed method above can be implemented as software, firmware, hardware and appropriate combinations thereof. Some physical components or all physical components can be implemented as software by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application specific integrated circuit. Such software can be distributed on a computer-readable medium, and the computer-readable medium can include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As known to those of ordinary skill in the art, the term computer storage medium is included in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data) and is volatile and non-volatile, removable and non-removable. Computer storage media includes but is not limited to RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassette, magnetic tape, disk storage or other magnetic storage device, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, as is well known to those skilled in the art, communication media typically embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0118] The above is a specific description of the preferred implementation of the present invention, but the invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A method for designing an AAV capsid protein, characterized in that: include: The AAV capsid protein sequence to be optimized is randomly initialized, and the initial temperature is set according to a preset temperature. The initial temperature represents the mutation possibility of each position of the initial AAV sequence; The cooling rate is determined based on the sum of the improvement rate of the targeting score and the yield score of the current AAV capsid protein sequence compared with the wild-type AAV capsid protein sequence, the diversity score of the current AAV capsid protein sequence, and the preset maximum cooling rate; Determining the current temperature based on the previous temperature and the cooling rate, selecting a first preset number of AAV capsid protein sequences at the current temperature based on an objective function, and continuing the cooling optimization until the AAV capsid protein sequences at the current temperature meet the preset index requirements, and then stopping the cooling search; the objective function includes a targeting score, a yield score, and a diversity score; Selecting a second preset number of undetermined AAV capsid protein sequences from all AAV capsid protein sequences that meet preset indicator requirements, clustering the undetermined AAV capsid protein sequences, and selecting a third preset number of AAV capsid protein sequences from each cluster of sequences as the final designed AAV capsid protein sequences; The diversity score of the AAV capsid protein sequence was obtained by the following steps: Obtaining a batch of AAV capsid protein sequences, and clustering the AAV capsid protein sequences according to the similarity and coverage between the AAV capsid protein sequences; The diversity score of each class of AAV capsid protein sequences is determined based on the total number of classes in the class; The diversity score of the AAV capsid protein sequence was calculated using the following formula: The cooling rate is determined based on the sum of the improvement rate of the targeting score and the yield score of the current AAV capsid protein sequence compared to the wild-type AAV capsid protein sequence, the diversity score of the current AAV capsid protein sequence, and the preset maximum cooling rate, specifically including: determining a first improvement rate based on a targeting score of the current AAV capsid protein sequence and a targeting score of a wild-type AAV capsid protein sequence; determining a second improvement rate based on the yield score of the current AAV capsid protein sequence and the yield score of the wild-type AAV capsid protein sequence; Determine an average value based on the first improvement rate and the second improvement rate, and calculate the ratio of the difference between 1 and the average value to the diversity score of the current AAV capsid protein sequence; The smaller value between the ratio and the preset maximum cooling rate probability is taken as the cooling rate.

2. The method for designing an AAV capsid protein according to claim 1, wherein: AAV capsid protein sequences were clustered based on their similarity and coverage, specifically including: Determine query and target sequences from a batch of AAV capsid protein sequences; The similarity is determined based on the ratio of the number of identical amino acids to the total number of amino acids in the query and target sequences; The coverage is determined based on the ratio of the length of the longest common subsequence between the query sequence and the target sequence to the total length of the query sequence; Sequences that meet both similarity and coverage requirements are clustered into the same category.

3. The method for designing an AAV capsid protein according to claim 1, wherein: The targeting score and yield score of the AAV capsid protein sequence were obtained by the following steps: The AAV capsid protein sequence was input into the trained estimation model to obtain the targeting score and yield score.

4. The method for designing an AAV capsid protein according to claim 1, wherein: According to the objective function, the first preset number of AAV capsid protein sequences at the current temperature are selected to continue the cooling optimization, specifically including: Mutate each position of the AAV capsid protein sequence according to the current temperature; Calculate the objective function value of all mutated AAV capsid protein sequences at the current temperature; Recalculating the cooling rate of the AAV capsid protein sequence and updating the current temperature of the AAV capsid protein sequence; The function values ​​are sorted by high to low, and the AAV capsid protein sequences with the first preset number of function value rankings are selected to continue the cooling optimization.

5. A design system for AAV capsid protein, characterized in that: include: The first module is used to randomly initialize the AAV capsid protein sequence to be optimized and set the initial temperature according to a preset temperature. The initial temperature represents the mutation possibility of each position of the initial AAV sequence; The second module is used to determine the cooling rate according to the sum of the improvement rate of the targeting score and the yield score of the current AAV capsid protein sequence compared with the wild-type AAV capsid protein sequence, the diversity score of the current AAV capsid protein sequence and the preset maximum cooling rate; The third module is used to determine the current temperature based on the temperature at the previous moment and the cooling rate, and select a first preset number of AAV capsid protein sequences at the current temperature according to the objective function to continue cooling optimization until the AAV capsid protein sequences at the current temperature meet the preset index requirements, and then stop the cooling search; the objective function includes the targeting score, the yield score and the diversity score; The fourth module is used to select a second preset number of undetermined AAV capsid protein sequences from all AAV capsid protein sequences that meet the preset indicator requirements, cluster the undetermined AAV capsid protein sequences, and select a third preset number of AAV capsid protein sequences from each class of sequences as the final designed AAV capsid protein sequences; The diversity score of the AAV capsid protein sequence was obtained by the following steps: Obtaining a batch of AAV capsid protein sequences, and clustering the AAV capsid protein sequences according to the similarity and coverage between the AAV capsid protein sequences; The diversity score of each class of AAV capsid protein sequences is determined based on the total number of classes in the class; The diversity score of the AAV capsid protein sequence was calculated using the following formula: The cooling rate is determined based on the sum of the improvement rate of the targeting score and the yield score of the current AAV capsid protein sequence compared to the wild-type AAV capsid protein sequence, the diversity score of the current AAV capsid protein sequence, and the preset maximum cooling rate, specifically including: determining a first improvement rate based on a targeting score of the current AAV capsid protein sequence and a targeting score of a wild-type AAV capsid protein sequence; determining a second improvement rate based on the yield score of the current AAV capsid protein sequence and the yield score of the wild-type AAV capsid protein sequence; Determine an average value based on the first improvement rate and the second improvement rate, and calculate the ratio of the difference between 1 and the average value to the diversity score of the current AAV capsid protein sequence; The smaller value between the ratio and the preset maximum cooling rate probability is taken as the cooling rate.

6. A design device for AAV capsid protein, characterized in that: include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the method according to any one of claims 1 to 4.

7. A computer-readable storage medium storing a program executable by a processor, characterized in that: The processor-executable program is used to perform the method according to any one of claims 1 to 4 when executed by the processor.

Images