A method for determining a protein structure alignment and related apparatus

Abstract

The application discloses a method for determining protein structure alignment and a related device. The method comprises the following steps: determining an initial quantum state to be prepared based on the structure of a target protein, wherein the target protein comprises a protein whose alignment mode is to be determined; constructing a target quantum circuit by using the initial quantum state and a Hamiltonian corresponding to the target protein; running and measuring the current target quantum circuit to obtain a final quantum state containing the alignment mode of the target protein; and determining the optimal alignment mode of the target protein according to the final quantum state. According to the embodiment of the application, the optimal alignment mode of the protein can be efficiently determined by using the eigen-parallelism of quantum computing.

Description

Technical Field

[0001] This application belongs to the field of quantum computing technology, and in particular to a method and related apparatus for determining protein structure alignment. Background Technology

[0002] Proteins, biological macromolecules formed by the combination of amino acids, perform many important functions in the life activities of organisms, such as immune responses, hormone regulation, and participation in and regulation of the cell life cycle. The number of amino acids varies greatly among different proteins, some containing around 50, while others can have over 2000. Research has found that proteins form complex three-dimensional structures (also known as natural structures) through the self-folding of amino acid sequences, and these natural structures are crucial for the correct fulfillment of life functions. If a protein forms an incorrect folded structure, the corresponding life activities will be affected, and in severe cases, it may even endanger the entire organism.

[0003] Scientists believe that protein function can be classified by its structure, and proteins with similar structures will have similar functions. If the structure of one protein is similar to that of another protein with a known function, then the function of that protein can also be reasonably predicted. Therefore, studying the similarity of protein structures, i.e., the protein structure alignment problem, is of great value to both basic and applied research in proteins and related fields. However, protein structure alignment of multiple sequences is an NP-hard (Non-deterministic Polynomial) problem, and the complexity of the object itself determines the difficulty of solving the problem. Furthermore, since most existing methods for solving the protein structure alignment problem are based on classical computation, the computational efficiency decreases significantly with the increase of amino acid sequences. Therefore, how to efficiently determine the optimal alignment of protein structures has become an urgent problem to be solved. Summary of the Invention

[0004] The purpose of this application is to provide a method and related apparatus for determining protein structure alignment, which aims to utilize the intrinsic parallelism of quantum computing to efficiently determine a better protein alignment.

[0005] One embodiment of this application provides a method for determining protein structure alignment, the method comprising:

[0006] Based on the structure of the target protein, the initial quantum state to be prepared is determined, wherein the target protein includes a protein whose alignment mode is to be determined;

[0007] Using the initial quantum state and the Hamiltonian corresponding to the target protein, a target quantum circuit is constructed;

[0008] Run and measure the current target quantum circuit to obtain the final quantum state containing the alignment of the target protein;

[0009] Based on the final quantum state, the optimal alignment of the target protein is determined.

[0010] Optionally, determining the initial quantum state to be prepared based on the structure of the target protein includes:

[0011] Two contact sets corresponding to the target protein are obtained, wherein each contact set contains at least one pair of contact elements, and the pair of contact elements consists of two amino acids in the target protein that satisfy a preset contact relationship;

[0012] The initial quantum state to be prepared is determined using two contact sets.

[0013] Optionally, determining the initial quantum state to be prepared using two contact sets includes:

[0014] Based on the two contact sets, the number of qubits required to prepare the initial quantum state is determined;

[0015] The determined qubits are grouped using a contact set;

[0016] For each group, determine the quantum state to be prepared;

[0017] Based on all the quantum states to be prepared, the initial quantum state is obtained.

[0018] Optionally, determining the number of qubits required to prepare the initial quantum state based on two contact sets includes:

[0019] Based on the number of the two contact sets, the number of qubits required to prepare the initial quantum state is determined by the following formula:

[0020] Q = (N+1)*M

[0021] Where Q is the required number of qubits, N is the number of contacts in one contact set, M is the number of contacts in another contact set, and N≥M;

[0022] The method of grouping the determined number of qubits using one of the contact sets includes:

[0023] The determined qubits are grouped into groups of N+1 qubits each.

[0024] Optionally, constructing the target quantum circuit using the initial quantum state and the Hamiltonian corresponding to the obtained target protein includes:

[0025] Construct an initial state preparation circuit for preparing the initial quantum state;

[0026] Construct a simulation circuit for simulating the evolution of the Hamiltonian corresponding to the target protein;

[0027] Construct adjustment circuits for adjusting the probability of quantum states;

[0028] By combining the initial state preparation circuit, the simulation circuit, and the adjustment circuit, the target quantum circuit is obtained.

[0029] Optionally, the Hamiltonian is represented by the following formula:

[0030]

[0031] Among them, H C Let q be the Hamiltonian. i q j K corresponds to the i-th and j-th qubits respectively. ij This is the penalty coefficient.

[0032] Optionally, the simulated circuit includes variational parameters;

[0033] The process of running and measuring the current target quantum circuit to obtain the final quantum state containing the alignment of the target protein includes:

[0034] Based on the parameter values ​​of the variational parameters in the current target quantum circuit, run and measure the current target quantum circuit to obtain the final state;

[0035] When the operation of the current target quantum circuit meets the preset termination condition, the current final state is taken as the final quantum state.

[0036] Another embodiment of this application provides a protein structure alignment determination device, the device comprising:

[0037] The first determining module is used to determine the initial quantum state to be prepared based on the structure of the target protein, wherein the target protein includes a protein whose alignment mode is to be determined;

[0038] A construction module is used to construct a target quantum circuit using the initial quantum state and the Hamiltonian corresponding to the obtained target protein;

[0039] The module is used to run and measure the current target quantum circuit to obtain the final quantum state containing the alignment of the target protein;

[0040] The second determining module is used to determine the optimal alignment of the target protein based on the final quantum state.

[0041] Optionally, the first determining module includes:

[0042] The obtaining unit is used to obtain two contact sets corresponding to the target protein, wherein the contact set contains at least one pair of contact elements, and the pair of contact elements consists of two amino acids in the target protein that satisfy a preset contact relationship;

[0043] A defining unit is used to determine the initial quantum state to be prepared using two contact sets.

[0044] Optionally, the determining unit is specifically used for:

[0045] Based on the two contact sets, the number of qubits required to prepare the initial quantum state is determined;

[0046] The determined qubits are grouped using a contact set;

[0047] For each group, determine the quantum state to be prepared;

[0048] Based on all the quantum states to be prepared, the initial quantum state is obtained.

[0049] Optionally, the determining unit is further specifically used for:

[0050] Based on the number of the two contact sets, the number of qubits required to prepare the initial quantum state is determined by the following formula:

[0051] Q = (N+1)*M

[0052] Where Q is the required number of qubits, N is the number of contacts in one contact set, M is the number of contacts in another contact set, and N≥M;

[0053] The determined qubits are grouped into groups of N+1 qubits each.

[0054] Optionally, the construction module is specifically used for:

[0055] Construct an initial state preparation circuit for preparing the initial quantum state;

[0056] Construct a simulation circuit for simulating the evolution of the Hamiltonian corresponding to the target protein;

[0057] Construct adjustment circuits for adjusting the probability of quantum states;

[0058] By combining the initial state preparation circuit, the simulation circuit, and the adjustment circuit, the target quantum circuit is obtained.

[0059] Optionally, the Hamiltonian is represented by the following formula:

[0060]

[0061] Among them, H C Let q be the Hamiltonian. i q j K corresponds to the i-th and j-th qubits respectively. ij This is the penalty coefficient.

[0062] Optionally, the simulated circuit includes variational parameters;

[0063] The obtaining module is specifically used for:

[0064] Based on the parameter values ​​of the variational parameters in the current target quantum circuit, run and measure the current target quantum circuit to obtain the final state;

[0065] When the operation of the current target quantum circuit meets the preset termination condition, the current final state is taken as the final quantum state.

[0066] One embodiment of this application provides a storage medium storing a computer program, wherein the computer program is configured to implement the method described in any of the above-described embodiments when running.

[0067] One embodiment of this application provides an electronic device including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to implement the method described in any of the above-described embodiments.

[0068] Compared with existing technologies, this application provides a method for determining protein structure alignment. First, based on the structure of the target protein, an initial quantum state to be prepared is determined, wherein the target protein includes proteins for which the alignment method is to be determined. Then, using the initial quantum state and the Hamiltonian corresponding to the obtained target protein, a target quantum circuit is constructed. Next, the current target quantum circuit is run and measured to obtain a final quantum state containing the alignment method of the target protein. Finally, based on the final quantum state, the optimal alignment method of the target protein is determined. Due to the superposition and entanglement of qubits, quantum computing possesses intrinsic parallelism. Therefore, utilizing the computational power of noisy qubits, the determination of protein structure alignment can be accelerated through quantum circuits, thereby efficiently determining a better protein structure alignment. Attached Figure Description

[0069] Figure 1 This is a network block diagram of a protein structure alignment determination system provided in an embodiment of this application;

[0070] Figure 2 A flowchart illustrating a method for determining protein structure alignment provided in an embodiment of this application;

[0071] Figure 3 A schematic diagram of a protein contact pattern provided in an embodiment of this application;

[0072] Figure 4 A schematic diagram illustrating a protein contact alignment method provided in an embodiment of this application;

[0073] Figure 5 A schematic diagram illustrating another protein contact alignment method provided in an embodiment of this application;

[0074] Figure 6 A schematic diagram of a tabular protein contact alignment method provided in an embodiment of this application;

[0075] Figure 7 This is a schematic diagram of the structure of a target quantum circuit provided in an embodiment of this application;

[0076] Figure 8 A schematic diagram of the equivalent quantum circuit structure of an entanglement module provided in an embodiment of this application;

[0077] Figure 9 This is a schematic diagram of a sub-circuit fabrication structure provided in an embodiment of this application;

[0078] Figure 10 This application provides a schematic diagram of the circuit structure after iSWAP gate decomposition in an embodiment of the present application;

[0079] Figure 11 This application provides a schematic diagram of the structure of an adjustment sub-circuit corresponding to an even number of qubits.

[0080] Figure 12 This application provides a schematic diagram of the structure of an adjustment sub-circuit corresponding to an odd number of qubits.

[0081] Figure 13 This is a schematic diagram of a protein structure alignment determination device provided in an embodiment of this application. Detailed Implementation

[0082] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application.

[0083] Figure 1This is a network block diagram of a protein structure alignment determination system provided in an embodiment of this application. The protein structure alignment determination system may include a network 110, a server 120, a wireless device 130, a client 140, a storage unit 150, a classical processing system 160, a quantum processing system 170, and may also include additional memory, a classical processor, a quantum processor, and other devices not shown.

[0084] Network 110 is a medium used to provide communication links between various devices and computers connected together within a system for determining protein structure alignment, including but not limited to the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof, and the connection method can be wired, wireless communication links, or fiber optic cables.

[0085] Server 120 and client 140 are conventional data processing systems that may contain data and applications or software tools that perform conventional computational processes. Client 140 may be a personal computer or a network computer, so the data may also be provided by server 120. Wireless device 130 may be a smartphone, tablet, laptop, smart wearable device, etc. Storage unit 150 may include database 151, which can be configured to store data such as qubit parameters, quantum logic gate parameters, quantum circuits, and quantum programs.

[0086] The classical processing system 160 (quantum processing system 170) may include a classical processor 161 (quantum processor 171) for processing classical data (quantum data) and a memory 163 (memory 172) for storing classical data (quantum data). The classical data (quantum data) may be a boot file, an operating system image, and an application program 162 (application program 173). The application program 162 (application program 173) may be used to implement a quantum algorithm compiled according to a protein structure alignment determination method provided in the embodiments of this application.

[0087] Any data or information stored or generated in the classical processing system 160 (quantum processing system 170) can also be configured to be stored or generated in another classical (quantum) processing system in a similar manner, and any application executed therein can also be configured to be executed in another classical (quantum) processing system in a similar manner.

[0088] It should be noted that a true quantum computer has a hybrid structure, which includes at least... Figure 1 The system consists of two main parts: the classical processing system 160, which is responsible for performing classical calculations and control; and the quantum processing system 170, which is responsible for running quantum programs and thus realizing quantum computing.

[0089] The aforementioned classical processing system 160 and quantum processing system 170 can be integrated into a single device or distributed across two different devices. For example, the first device, including the classical processing system 160, runs a classical computer operating system that provides quantum application development tools and services, as well as the storage and network services required for quantum applications. Users develop quantum applications using the quantum application development tools and services on the second device and send the quantum program to the second device, including the quantum processing system 170, via the network services. The second device runs a quantum computer operating system, which parses the code of the quantum program and compiles it into instructions that can be recognized and executed by the quantum computer control system. The quantum processor 170 then implements the quantum algorithm corresponding to the quantum program based on these instructions.

[0090] In the classic silicon-based processing system 160, the units of the classic processor 161 are CMOS transistors. These computing units are not limited by time or coherence; that is, they are available at any time without time constraints. Furthermore, the number of these computing units in a silicon chip is sufficient; currently, a classic processor contains tens of thousands of computing units. The sufficient number of computing units and the fixed selectable computing logic of the CMOS transistors, such as AND logic, allow for computational efficiency through a combination of numerous CMOS transistors and limited logic functions.

[0091] Unlike the logic units in the classical processing system 160, the basic computational unit of the quantum processor 171 in the quantum processing system 170 is the qubit. The input of a qubit is limited by coherence and coherence time; that is, a qubit is limited by its available usage time and is not always readily available. Making full use of qubits within their available usage time is a key challenge in quantum computing. Furthermore, the number of qubits in a quantum computer is one of the representative indicators of its performance. Each qubit performs computational functions through on-demand configured logic functions. Given the limited number of qubits and the diverse logic functions available in quantum computing, such as Hadamard gates (H gates), Pauli-X gates (X gates), Pauli-Y gates (Y gates), Pauli-Z gates (Z gates), X gates, RY gates, RZ gates, CNOT gates, CR gates, iSWAP gates, Tofoli gates, etc., quantum computing requires combining a limited number of qubits with diverse logic function combinations to achieve computational effects.

[0092] Based on these differences, the design of logical functions applied to qubits (including the design of whether qubits are used and the design of the efficiency of each qubit's use) is crucial to improving the computational performance of quantum computers and requires specialized design. The aforementioned design considerations for qubits are technical problems that ordinary computing devices do not need to address. Therefore, this application proposes a method and related apparatus for determining protein structure alignment in quantum computing, aiming to efficiently determine the optimal protein structure alignment by utilizing the intrinsic parallelism of quantum computing.

[0093] The following is a brief introduction to the applications of protein structure alignment in the field of biochemistry:

[0094] Protein structure alignment is the process of determining the structural similarity between two proteins whose structural information is already known. Research on protein structure alignment can help solve problems including, but not limited to, the following:

[0095] Given a newly determined three-dimensional protein structure, find proteins with similar structures in a structure database to obtain the structurally conserved regions of unknown proteins and infer their evolutionary information or predict their functions.

[0096] By constructing a classification database of proteins with known structures through structural alignment, we can analyze the functions of different proteins and their corresponding structural features, and use this information to explore the mechanisms by which proteins achieve their functions.

[0097] Based on known protein structures, possible protein folding patterns can be summarized to help predict protein structures, i.e., evaluate the predicted structures, and lay the foundation for further protein molecule design.

[0098] Therefore, solving the protein structure alignment problem and determining the optimal protein alignment method can promote the development of the biological, medical and pharmaceutical fields, such as understanding the genetic evolution mechanism of species, treating rare diseases, and designing small molecule drugs.

[0099] See Figure 2 , Figure 2 A flowchart illustrating a method for determining protein structure alignment provided in this application embodiment may include the following steps:

[0100] S201: Based on the structure of the target protein, determine the initial quantum state to be prepared, wherein the target protein includes proteins whose alignment needs to be determined.

[0101] The target protein comprises two proteins whose alignment needs to be determined. The structure of the target protein can be manually input, obtained through other devices, or stored in a database. The initial quantum state may contain the alignment of the target protein. The initial quantum state can be constructed using the amino acid sequence of the target protein. For example, the amino acid sequence can be processed into its corresponding binary form, and then the binary amino acid sequence can be processed to determine the initial quantum state. Alternatively, the target protein can be constructed by determining all alignments based on its corresponding structure, encoding all alignments to obtain the initial quantum state, or by other methods, which are not listed here.

[0102] S202: Construct the target quantum circuit using the initial quantum state and the Hamiltonian corresponding to the obtained target protein.

[0103] The Hamiltonian is the Hamiltonian operator corresponding to the protein structure alignment problem, and it is the mathematical representation of the protein structure alignment method. The target quantum circuit contains quantum logic gates or combinations of quantum logic gates that implement the corresponding function. Specifically, it includes the function of preparing the initial quantum state and the function of simulating the Hamiltonian. Because one quantum logic gate can be equivalent to other quantum logic gates or combinations of quantum logic gates, and different quantum computers may support different types of quantum logic gates, the structure of the target quantum circuit may differ, but the function implemented is the same.

[0104] S203: Run and measure the current target quantum circuit to obtain the final quantum state containing the alignment of the target protein.

[0105] The current target quantum circuit is the latest version. While the target quantum circuit may change slightly with iterative operation, its overall structure will remain similar. After preparing the initial quantum state, the initial quantum state is applied to the Hamiltonian to simulate its evolution. Measurements are then performed on the target quantum circuit to obtain the final state. When the operation of the target quantum circuit satisfies the preset convergence condition, the current final state of the target quantum circuit is the final quantum state.

[0106] S204: Determine the optimal alignment of the target protein based on the final quantum state.

[0107] The final quantum state may have multiple quantum states, each corresponding to a specific alignment. The optimal alignment can be the one corresponding to the quantum state with the highest probability, or the one corresponding to the quantum state with the lowest expected energy. It's important to note that the determined optimal alignment may not be the globally optimal alignment for the target protein structure. However, a sufficiently good alignment is one that closely approximates the globally optimal alignment when the target quantum circuit meets the convergence condition; the degree of closeness depends on the convergence condition. Once the optimal alignment is determined, the similarity of the corresponding proteins is also determined. This method, leveraging the intrinsic parallelism of quantum computing, can rapidly solve the protein structure alignment problem, thereby improving the efficiency of determining the most similar protein structure and promoting advancements in biochemistry.

[0108] As can be seen, the embodiments of this application first determine the initial quantum state to be prepared based on the structure of the target protein; then, using the initial quantum state and the Hamiltonian corresponding to the obtained target protein, a target quantum circuit is constructed; then, the current target quantum circuit is run and measured to obtain a final quantum state containing the alignment of the target protein; finally, based on the final quantum state, the optimal alignment of the target protein is determined. Due to the superposition and entanglement of qubits, quantum computing has intrinsic parallelism. Therefore, by utilizing the computing power of noisy qubits, the alignment of protein structures can be determined more quickly through quantum circuits, thereby efficiently determining a better alignment of protein structures.

[0109] In some possible embodiments of this application, determining the initial quantum state to be prepared based on the structure of the target protein may include:

[0110] Two contact sets corresponding to the target protein are obtained, wherein each contact set contains at least one pair of contact elements, and the pair of contact elements consists of two amino acids in the target protein that satisfy a preset contact relationship;

[0111] The initial quantum state to be prepared is determined using two contact sets.

[0112] Protein folding is achieved through the binding of amino acids; therefore, the structural information of any protein can be described by the binding pattern of its constituent amino acids. From a physical conformational perspective, the binding pattern of amino acids can be viewed both as information about the important chemical bonds connecting the amino acids and as information about the relative spatial distances between them. If two amino acids meet a preset contact condition—which could be that the important chemical bonds between the two amino acids are less than a certain threshold (the specific size can be set manually, typically on the order of 0.1 nanometers)—these two amino acids are considered to be in contact. In constructing a protein contact map, the two amino acids that are closest in sequence number are ignored; therefore, amino acids whose sequence numbers are nearest neighbors will never be connected on the contact map. Figure 3 This diagram illustrates the contact graph of a protein containing eight amino acids. Each node represents an amino acid and is arranged in the protein's sequence. Two nodes connected by a solid line indicate that they are in physical contact. The contact set for this protein includes five elements: A1-A3, A1-A6, A2-A6, A3-A6, and A6-A8. This contact set can be received from other devices or obtained by processing the protein's structural information. Specifically, it is obtained by constructing a contact graph based on the protein's structural information and then traversing that graph.

[0113] Once the two contact sets corresponding to the target protein are determined, the possible alignment methods between the proteins are determined based on each element in the two contact sets. All alignment methods are processed to obtain the initial quantum state, which can contain all alignment methods.

[0114] In some possible embodiments of this application, determining the initial quantum state to be prepared using two contact sets may include:

[0115] Based on the two contact sets, the qubits required to prepare the initial quantum state are determined;

[0116] The determined qubits are grouped using a contact set;

[0117] For each group, determine the quantum state to be prepared;

[0118] Based on all the quantum states to be prepared, the initial quantum state is obtained.

[0119] The number of qubits required to prepare the initial quantum state can be the product of the number of contacts in two contact sets. For example, if one contact set includes 3 contacts and the other contains 4 contacts, then the number of qubits is 12. Alternatively, the number of qubits can be determined based on the number of contacts in the two contact sets. Once the number of qubits is determined, the number of qubits is also determined. The qubits can be grouped according to the number of contacts in one of the contact sets, with one contact corresponding to one group. Each quantum state to be prepared includes all alignments corresponding to that contact. Combining the quantum states to be prepared in each group yields the initial quantum state.

[0120] In some possible embodiments of this application, determining the number of qubits required to prepare the initial quantum state based on two contact sets may include:

[0121] Based on two contact sets, the number of qubits required to prepare the initial quantum state is determined by the following formula:

[0122] Q = (N+1)*M

[0123] Where Q is the required number of qubits, N is the number of contacts in one contact set, M is the number of contacts in another contact set, and N≥M;

[0124] The process of grouping the determined number of qubits using one of the contact sets may include:

[0125] The determined qubits are grouped into groups of N+1 qubits each.

[0126] For any two proteins, we can choose a set of mappings, i.e., an alignment. This set of mappings is a pairwise combination of the nodes of the contact graphs of the two proteins. Figure 4 and Figure 5 Two mapping methods, i.e. two alignment methods, are shown respectively. Figure 4 The alignment is valid. Figure 5 Alignment is invalid. A valid alignment must satisfy the following conditions: the node mapping is one-to-one; the node mapping order must be monotonic. If we construct a contact graph of two proteins based on their sequence order and draw the corresponding node combinations with dashed lines, then a valid alignment on the contact graph is represented by the following rules: Rule 1: No two dashed lines appear on any node; Rule 2: No dashed lines intersect.

[0127] Figure 5This is an illegal alignment. Node B1 emits two dashed lines, violating rule 1. The two dashed lines A3B5-A6B3 intersect, violating rule 2. For the purpose of aligning structures, rule 1 is more natural. Rule 2 is established for the following reason: if we want to align the similarity of the folded structures of two proteins, then considering the positions of the remaining amino acids bound to amino acids in sequence order is a reasonable method. Since it is based on sequence order, establishing rule 2 is logically natural.

[0128] Suppose we are considering two proteins and choosing a (legal) alignment. Figure 4 Taking legal alignment as an example, if a contact in one protein corresponds to two nodes located within the selected alignment, such as the contact formed by A1 and A6, and the corresponding alignments of these two nodes in another protein are B1 and B5, it can be observed that these four nodes (A1, A6, B5, B1) form a loop in the contact graph. Such a loop can be considered an overlap of the contact graphs of the two proteins. Contact graph-based protein structure alignment involves finding one or more (legal) alignments with the largest overlap in the contact graphs of two proteins.

[0129] Besides using contact diagrams to represent the alignment of two proteins, tables or other similar formats, such as a chessboard, can also be used. Representing them in a table or similar format is more suitable for mathematical and computational processing. The table can be two-dimensional, with each row corresponding to one contact of the first protein and each column corresponding to one contact of the second protein. For example... Figure 6 As shown, in Figure 6 The table uses an ascending order of two nodes to represent an alignment. Any k unique points in the table can represent an alignment (even if it's invalid). However, any single point in the table always represents a valid alignment, contributing one unit of overlap. Adding another point to the table increases the overlap by one if the overall alignment remains valid. This continues until the number of points (k) cannot be increased further, at which point an alignment with overlap of k is found. The similarity between the two protein structures among all possible alignments represents the overlap of the alignment with the highest overlap. Essentially, this means finding a pairwise valid set of points in the table that maximizes the total overlap—that is, finding the subset of points with the largest number of elements.

[0130] For any pair of proteins A and B, there is a one-to-one correspondence between the optimal solution of a maximal clique in the table and the protein alignment problem. Furthermore, the maximal clique problem and the maximal independent subset problem are conjugates and can be transformed into each other in constant time. Therefore, given a pair of proteins, we first identify all their contacts and organize them in a table, listing all nodes and constraints, thus transforming it into a maximal independent subset problem. This problem can then be solved using a quantum combinatorial optimization algorithm.

[0131] Figure 6 The table hints at the hidden structure of the maximum independent subset problem. Since protein structure alignment requires that a node cannot have multiple alignments, each row or column can only select one alignment. Therefore, an empty column can be added to the table to represent an empty set contact; it does not represent any contact. The column containing an empty set contact does not represent any alignment, but rather the contact that is not selected in the corresponding row. Based on this, the table has two characteristics: First, if an alignment is not located in any position in an empty column, it and its corresponding alignment in the same row and column are necessarily mutually exclusive. This is because a node cannot have multiple pairings, so each row or column can only select one pairing. For an alignment located in any position in an empty column, it and its corresponding alignment in the same row are also necessarily mutually exclusive. This is because contacts in the same row either select a contact to form an alignment, or are not selected in an empty column. Since not selecting multiple alignments corresponding to the contact set does not affect the validity of the alignment, two positions in an empty column are not mutually exclusive.

[0132] When the contact set with more contacts in two contact sets is used as the column element, and the contact set with fewer contacts in two contact sets is used as the row element, adding an empty column can reduce the number of qubits required. In the embodiments of this application, the number of qubits can be (N+1)*M.

[0133] For a table with M rows and N+1 columns, use (N+1)*M qubits to encode the alignment one-to-one according to the order of the first row being the 0th to the Nth qubit, the second row being the N+1th to the 2Nth qubit, and so on. Construct the quantum state to be prepared row by row. The quantum state to be prepared can be a W state.

[0134] In some possible embodiments of this application, constructing the target quantum circuit using the initial quantum state and the Hamiltonian corresponding to the target protein includes:

[0135] Construct an initial state preparation circuit for preparing the initial quantum state;

[0136] Construct a simulation circuit for simulating the evolution of the Hamiltonian corresponding to the target protein;

[0137] Construct adjustment circuits for adjusting the probability of quantum states;

[0138] By combining the initial state preparation circuit, the simulation circuit, and the adjustment circuit, the target quantum circuit is obtained.

[0139] The quantum logic gates contained in the initial state preparation circuit can be determined by the initial quantum state to be prepared. For example, if the initial quantum state is an equal-probability superposition state, then the initial state preparation circuit is constructed based on applying H gates to each quantum bit. The Hamiltonian can be expanded into the sum of multiple sub-terms. By simulating each sub-term, a simulation circuit can be obtained. The role of the simulation circuit is to evaluate the quality (loss function) of the solution corresponding to the encoded quantum state, and the result is reflected in the phase change of the quantum state. The phase change of each quantum state is related to the magnitude of its corresponding loss function, so it can also be called a phase separation circuit. The loss function is related to the energy of the final state of the quantum circuit, and can be the energy expectation, the Gibbs function, the energy expectation of CVaR (Conditional Value at Risk) sampling, Fisher information, or other functions. After the simulation circuit, although the quality of the solution corresponding to each quantum state is reflected in its phase, the phase cannot be directly measured. Therefore, the circuit needs to be adjusted and applied again. With the help of the simulation circuit, the probability of a higher-energy state can be increased, realizing the state transition. The circuit adjustment can be constructed in different ways depending on the situation. It should be noted that the simulation circuit contains at least one variational parameter, and the adjustment circuit also contains at least one variational parameter. In the embodiments of this application, the initial quantum state can be determined by the adjustment circuit. There is a correlation between the structure of the adjustment circuit and the initial quantum state. When the structure of the adjustment circuit is determined, the initial quantum state is determined based on this correlation. When the initial quantum state, the structure of the adjustment circuit, and the Hamiltonian are determined, the overall structure of the target quantum circuit is also determined accordingly.

[0140] The entire target quantum circuit contains at least one analog circuit and at least one adjustment circuit. The analog circuit and the adjustment circuit alternate, and one analog circuit and one adjustment circuit can be considered as one layer. A target quantum circuit can contain multiple layers, and the number of layers can be given arbitrarily. For example, taking 4 qubits as an example, the target quantum circuit can be as follows: Figure 7As shown, the initial state of the qubit is set to |0>. The number of repetitions p of the simulation circuit and the adjustment circuit is given manually. Depending on the specific design, each layer of the simulation circuit contains at least one variational parameter γ, and the adjustment circuit also contains at least one variational parameter β. These variational parameters are related to the rotation angle of the rotating gate in the circuit. The variational parameters of each layer can be unequal, and the number of parameters contained in each layer can also be unequal. The rotation angle of the rotating gate of the simulation circuit can be determined by the corresponding variational parameter and the coefficient of the Hamiltonian. For example, the rotation angle = 2 * the corresponding variational parameter * the coefficient of the Hamiltonian. The rotation angle of the rotating gate of the simulation circuit can be determined by the corresponding variational parameter and the preset adjustment multiple. For example, for a p-layer target quantum circuit, the variational parameters of the first layer simulated circuit can be γ1 = (γ11, γ12, γ13), the variational parameters of the first layer adjusted circuit can be β1 = (β11, β12, β13, β14), the variational simulation parameters of the second layer can be γ2 = (γ21, γ22), the variational parameters of the second layer adjusted circuit can be β2 = (β21, β22, β23, β24, β25, β26), and so on.

[0141] In some possible embodiments of this application, the construction of the initial state preparation circuit for preparing the initial quantum state may include:

[0142] For each group of qubits, perform the following operations:

[0143] Each pair of adjacent numbered qubits is grouped into a first subgroup;

[0144] An entanglement module and a SWAP gate are applied to each first subgroup, and a Pauli-X gate is applied to the qubit with the largest number, so as to obtain the preparation sub-circuit for constructing the quantum state to be prepared corresponding to the group.

[0145] By combining all the preparation sub-circuits, the initial state preparation circuit for preparing the initial quantum state is obtained.

[0146] The specific form of the quantum state to be prepared can be determined by adjusting the circuit structure. When the circuit includes an XY mixer, the quantum state to be prepared can be a W state. The W state is defined as an equiprobable superposition of "single excited states", mathematically written as:

[0147]

[0148] The W state is an equiprobable superposition of states with a spin number of 1. It is an equiprobable superposition of all states with a Hamming weight of 1 under the computational basis and has the characteristic of spin number conservation under the action of the XY mixer.

[0149] The entanglement module can include Pauli-X gates, CNOT gates, and RY gates. This rotating gate contains parameters related to the number of qubits in the target quantum circuit and the qubit numbering of the entanglement module. The values ​​of these parameters satisfy the following conditions: j represents the bit number currently being acted upon by the entanglement module (i.e., the bit acted upon by the RY gate). An entanglement module can be represented as U(θ), ​​and its equivalent quantum circuit can be represented as follows: Figure 8 As shown, an entangled module includes two CNOT gates, two Pauli-X gates, and one RY gate. The θ value in each entangled module varies depending on the active qubit. It should be noted that the parameter values ​​in the initial state preparation circuit are fixed, while the variational parameters in the target quantum circuit refer to the parameters contained in the simulation and adjustment circuits.

[0150] For example, a qubit group is q0, q1, q2, q3, q4. Adjacent qubits are grouped together to obtain groups of (q0, q1), (q1, q2), (q2, q3), and (q3, q4). An entanglement module and a SWAP gate are applied to each group, and finally, a Pauli-X gate is applied to q4. This yields... Figure 9 The fabrication sub-circuit shown can include M parallel fabrication sub-circuits in its initial state.

[0151] Because the topology of a real chip is linear, meaning that each qubit can only interact with its nearest neighbor, using adjacent qubits as a group allows the fabricated sub-circuits to be mapped onto the actual chip.

[0152] In some possible embodiments of this application, the construction of the initial state preparation circuit for preparing the initial quantum state includes:

[0153] For each group of qubits, perform the following operations:

[0154] Assign an auxiliary bit to the group of qubits;

[0155] Each auxiliary bit is grouped together with each qubit in the group of qubits to form a second group.

[0156] An entanglement module is applied to each second subgroup, and a Pauli-X gate is applied to the auxiliary qubit to obtain a preparation subcircuit for constructing the quantum state to be prepared corresponding to the group.

[0157] By combining all the preparation sub-circuits, the initial state preparation circuit for preparing the initial quantum state is obtained.

[0158] In this embodiment, the W-state is constructed using a sequential generation method. This method requires an additional auxiliary qubit and is characterized by sequentially interacting the data bits (the bits used for encoding) with the auxiliary bit; hence, it is called the sequential generation method. The sequential generation method can be conveniently applied to quantum chips with linear topologies.

[0159] For example, if a group of qubits contains d qubits and their corresponding auxiliary bits, all in the 0 state, a parametric quantum circuit module U(θ) is constructed as follows. First, the entanglement module is applied to the 0th qubit and the auxiliary bit. At this point, the parameters of the entanglement module are... Then this entanglement module is applied to bit 1 and the auxiliary bit, at which point the entanglement module parameters are... This process continues until all bits are entangled with the auxiliary bits. At this point, the auxiliary bits must be in a 1 state. Then, an X-gate is applied to the auxiliary bits to flip them. After the above steps, the qubit group is in a W state, and the auxiliary bits are not entangled with other bits and are in a 0 state, which can be directly used to construct the required W state.

[0160] In some possible embodiments of this application, the construction of the adjustment circuit for adjusting the quantum state probability includes:

[0161] For each group of qubits, the corresponding third subgroup is obtained based on the number and number of qubits in that group;

[0162] By applying adjustment modules to the qubits of each group of three subgroups, adjustment sub-circuits corresponding to each third subgroup are obtained.

[0163] Combine all the corresponding adjustment sub-circuits of the subgroups to obtain the adjustment circuit unit;

[0164] Combine all the adjustment circuit units to obtain the adjustment circuit used to adjust the probability corresponding to the quantum state.

[0165] In quantum computing, the basic unit of information is the qubit. A qubit has two states, 0 and 1, denoted as |0> and |1>. However, it can exist in a superposition of these two states, which can be represented as |ψ> = a|0> + b|1>, where a and b are complex numbers representing the amplitudes (probability amplitudes) of the |0> and |1> states, respectively. This is not possible with classical bits. After measurement, the state of a qubit collapses to a definite state (eigenstate, here |0> or |1>), where the probability of collapsing to |0> is |a|. 2 The probability of collapsing to |1> is |b|. 2 , |a| 2 +|b| 2=1, |> represents the Dirac notation. In this embodiment, the adjustment module mainly further filters the solutions of the quantum state by adjusting the probabilities mentioned above. Specifically, if the quantum state to be prepared is an equal-probability superposition state, the adjustment module only contains the RX gate, i.e., the X mixer. If the quantum state to be prepared is the W state, the adjustment circuit contains the XY mixer. The purpose of using the XY mixer is to utilize the characteristic that contacts in the same contact set corresponding to the same group are mutually repulsive, in conjunction with the W state, to ensure that the total Hamming weight remains unchanged during the evolution of the quantum state, thereby limiting the size of the search solution space and excluding illegal solutions corresponding to multiple alignment methods for the same contact. The target quantum circuit will not process these excluded solutions, saving computational resources.

[0166] In this embodiment, a corresponding adjustment sub-circuit is constructed for each group of qubits. According to a preset grouping method, a corresponding third subgroup is obtained. The adjustment module is applied to each third subgroup to obtain the adjustment sub-circuit corresponding to each third subgroup. That is, each subgroup corresponds to an XY mixer. Then, these adjustment sub-circuits are combined in a certain order to obtain the adjustment circuit.

[0167] The XY mixer contains multiple units, each of which is a module operating on two bits. This module is the adjustment module, and its Hamiltonian form is:

[0168] H M =X1X2+Y1Y2

[0169] When a two-qubit system is in the W state, this Hamiltonian guarantees the conservation of the system's spin number, equivalent to a SWAP gate. Since the adjustment circuit is parametric, it can be simulated using an iSWAP gate. Furthermore, an iSWAP can also be decomposed using basic quantum logic gates, such as... Figure 10 The route shown.

[0170] The entire XY mixer circuit consists of multiple units as described above, each acting on a different qubit. Therefore, the XY mixer circuit can be represented by pairwise groups of all qubits. The grouping method and even the order of grouping have a certain impact on the number of layers and the effect of the circuit. An important purpose of grouping is to minimize the number of layers in the circuit, in other words, to maximize the number of units that can act simultaneously; at the same time, it is also necessary to ensure that qubits at different positions have the opportunity to swap with each other. Based on this, different methods can be used to group the qubits, but the grouping method must ensure spin number conservation.

[0171] In some possible embodiments of this application, obtaining a corresponding third subgroup for each group of qubits, based on the number and number of qubits in that group, includes:

[0172] Each qubit is grouped, and based on the parity of the determined number of qubits, the corresponding qubits are grouped in pairs according to the nearest neighbor parity principle to obtain the corresponding third subgroup.

[0173] Alternatively, for each group of qubits, obtain the binary number corresponding to the qubit number contained in that group; based on the obtained binary number and the preset grouping method, group the qubits corresponding to the binary number to obtain the corresponding third subgroup.

[0174] This application provides two grouping methods. The first method uses simple nearest-neighbor parity grouping, and the corresponding adjustment sub-circuit can also be called a parity mixer. Using this method, the parity mixer only needs two layers, but the qubit swapping is relatively limited. The entire parity mixer allows at most one bit to interact with its nearest-neighbor bit. However, its advantage is a simpler circuit topology and fewer layers. The grouping method of the parity separation mixer is affected by the parity of the number of qubits corresponding to the overlapping region. When the number of qubits is even, the qubits can be directly grouped. For example, such as... Figure 11 As shown, taking a 6-qubit qubit example, the qubits are grouped as (q0, q1), (q1, q2), (q2, q3), (q3, q4), (q4, q5), and (q5, q0). Adjustment modules are then applied to (q0, q1), (q2, q3), and (q4, q5) as the first layer of the mixer, and to (q1, q2), (q3, q4), and (q5, q0) as the second layer. It should be noted that since the adjustment modules applied to q0 and q5 cannot be directly represented on the diagram, different colors are used to indicate their application to q0 and q5. Assuming the mixer rotation angle is π, this grouping method shifts each even-numbered bit down by 2 bits and each odd-numbered bit up by 2 bits, while maintaining periodic boundary conditions and forming a simple ring structure.

[0175] When the number of qubits corresponding to a group is odd, an additional auxiliary bit needs to be introduced to make the number of qubits even. This is because when the number of qubits is odd, it cannot be guaranteed that each layer of the constructed mixer will have the same structure. Figure 12It can be seen that when the number of qubits is 5, the qubits are grouped according to odd and even values ​​as (q0, q1), (q1, q2), (q2, q3), and (q3, q4). Adjustment modules are then applied to (q0, q1) and (q2, q3) as the first layer of the mixer, and to (q1, q2) and (q3, q4) as the second layer. The structures of the first and second layers are different. In this embodiment, an auxiliary bit can be introduced. This auxiliary bit only participates in the initial state preparation and adjustment circuits, not in the simulation circuits, ensuring that the total number of qubits is even without affecting the validity of the solution.

[0176] Another grouping method can be based on binary encoding classification. If the number of qubits corresponding to the overlapping region is 4, first convert the qubit numbers contained in the W state (starting from 0) into binary numbers b1b2. For example, bit 0 should be written as 00, and bit 3 should be written as 11. Define operation O. b ({i}), its function is to flip the value of the binary number corresponding to the set {i}, for example, O({i}). b ({1, 2}) represents flipping b1 and b2. Each operation defines a grouping method, for example, O b ({1}) will flip the position of b1, so 0b2 and 1b2 will be grouped together, that is, 00-10 (bits 0 and 2) in one group, and 01-11 (bits 1 and 3) in another group. Similarly, there are two other operations O b ({1, 2}) and O b ({2}), their grouping methods are as follows: 00-11 group (bit 0 and bit 3), 01-10 group (bit 1 and bit 2); 00-01 group (bit 0 and bit 1), 10-11 group (bit 2 and bit 3). Due to the exclusivity of the swap, operation O... b Units in different groups defined by ({i}) will not act on the same bit, and therefore belong to the same layer when expressed as a quantum circuit. Excluding operations that do not flip any bits, there are only n-1 operations in total, so the number of layers in the circuit is n-1. Mathematically, it can be proven that when the number of qubits is exactly a power of 2, the W state is an eigenstate of the circuit constructed in the above manner to adjust its eigenstates.

[0177] In some possible embodiments of this application, the Hamiltonian can be expressed by the following formula:

[0178]

[0179] Among them, H C Let q be the Hamiltonian. i q jK corresponds to the i-th and j-th qubits respectively. ij This is the penalty coefficient.

[0180] q i q j In simulation, this can be represented using the Pauli Z-gate as (IZ) / 2. The first term of the Hamiltonian is the single-body action term, which sums all alignments formed by two contact sets. If the two contact sets represent expectations, the first single-body action term sums all cells on the chessboard, reflecting the total number of selected alignments, i.e., the total number of contact overlaps. The second term is the two-body action term, specifically the penalty term, which sums all alignments that are not on the same row. For example, when the selected q... i q j If the two corresponding contacts are illegally aligned, then the corresponding K ij =2, otherwise 0, K ij The specific value can be set according to the actual situation. The penalty term makes the energy of illegal alignment higher, so that the legal maximum independent set becomes the ground state described by the Hamiltonian.

[0181] In some possible embodiments of this application, the step of running and measuring the current target quantum circuit to obtain a final quantum state containing the alignment of the target protein includes:

[0182] Based on the parameter values ​​of the variational parameters in the current target quantum circuit, run and measure the current target quantum circuit to obtain the final state;

[0183] When the operation of the current target quantum circuit meets the preset termination condition, the current final state is taken as the final quantum state.

[0184] When initial parameter values ​​are obtained, various methods can be employed. If the target quantum circuit has only one layer, a scanning method can be used. Specifically, within the obtained variational parameter range, a series of parameter values ​​are obtained with a preset step size. These values ​​are then used as parameters in the target quantum circuit, and the corresponding target quantum circuits are run to measure the final state. The quantum state with the highest probability is then obtained, and the parameter value corresponding to the minimum function value is selected as the initial parameter value. If the target quantum circuit contains multiple layers, but the number is small, a random method can be used to obtain the initial parameter values. Specifically, within the obtained parameter range, some parameter values ​​are randomly selected and used as variational parameter values ​​for the target quantum circuit. The corresponding loss function value is obtained, and the initial parameter value is determined based on the loss function value. Initial parameter values ​​can also be obtained by fine-tuning parameter values ​​obtained from solving other combinatorial optimization problems, or they can be obtained through other methods.

[0185] When the target quantum circuit contains a relatively large number of layers, the initial parameter values ​​obtained using the methods described above can be used to predict the initial parameter values ​​of other layers. These prediction methods can include linear interpolation, Fourier interpolation, or random guessing based on the initial parameter values, etc. It should be noted that if the target quantum circuit contains more than one variational parameter, the initial parameter values ​​for each variational parameter can be determined separately using the methods described above.

[0186] Whether to terminate the iterative operation of the target quantum circuit depends on a preset termination condition. The termination condition can be that the current iteration count reaches the maximum iteration count, the loss function value is within a preset precision, the loss function value converges to its minimum, or the number of times a quantum state with a probability greater than a preset threshold occurs in the final state exceeds a preset number. When the termination condition is met, the target quantum circuit terminates its iteration.

[0187] When the convergence condition is not met, the values ​​of the variational parameters in the target quantum circuit need to be updated. Specifically, an optimizer can be used to update the variational parameters. The optimizer can optimize the variational parameters using the final state. The specific process of this step depends on the classical optimizer used. Classical optimizers can employ gradient optimization methods, non-gradient optimization methods, machine learning methods, etc. If there is more than one variational parameter, the corresponding parameter values ​​are updated separately. When the convergence condition is met, it means that the final state is the desired final quantum state. When the target quantum circuit contains more than one variational parameter, the above method can be used to update the parameter values ​​of these variational parameters separately.

[0188] See Figure 13 , Figure 13 This is a schematic diagram of a protein structure alignment determination device provided in an embodiment of this application. Figure 2 Corresponding to the process shown, the apparatus includes:

[0189] The first determining module 1301 is used to determine the initial quantum state to be prepared based on the structure of the target protein, wherein the target protein includes a protein whose alignment mode is to be determined;

[0190] Construction module 1302 is used to construct a target quantum circuit using the initial quantum state and the Hamiltonian corresponding to the obtained target protein;

[0191] The module 1303 is used to run and measure the current target quantum circuit to obtain the final quantum state containing the alignment of the target protein;

[0192] The second determining module 1304 is used to determine the optimal alignment of the target protein based on the final quantum state.

[0193] In some possible implementations of this application, the first determining module 1301 may include:

[0194] The obtaining unit is used to obtain two contact sets corresponding to the target protein, wherein the contact set contains at least one pair of contact elements, and the pair of contact elements consists of two amino acids in the target protein that satisfy a preset contact relationship;

[0195] A defining unit is used to determine the initial quantum state to be prepared using two contact sets.

[0196] In some possible embodiments of this application, the determining unit may be specifically used for:

[0197] Based on the two contact sets, the number of qubits required to prepare the initial quantum state is determined;

[0198] The determined qubits are grouped using a contact set;

[0199] For each group, determine the quantum state to be prepared;

[0200] Based on all the quantum states to be prepared, the initial quantum state is obtained.

[0201] In some possible embodiments of this application, the determining unit may also be specifically used for:

[0202] Based on the number of the two contact sets, the number of qubits required to prepare the initial quantum state is determined by the following formula:

[0203] Q = (N+1)*M

[0204] Where Q is the required number of qubits, N is the number of contacts in one contact set, M is the number of contacts in another contact set, and N≥M;

[0205] The determined qubits are grouped into groups of N+1 qubits each.

[0206] In some possible embodiments of this application, the construction module 1302 may be specifically used for:

[0207] Construct an initial state preparation circuit for preparing the initial quantum state;

[0208] Construct a simulation circuit for simulating the evolution of the Hamiltonian corresponding to the target protein;

[0209] Construct adjustment circuits for adjusting the probability of quantum states;

[0210] By combining the initial state preparation circuit, the simulation circuit, and the adjustment circuit, the target quantum circuit is obtained.

[0211] In some possible embodiments of this application, the Hamiltonian can be expressed by the following formula:

[0212]

[0213] Among them, H C Let q be the Hamiltonian. i q j K corresponds to the i-th and j-th qubits respectively. ij This is the penalty coefficient.

[0214] In some possible embodiments of this application, the analog circuit includes variational parameters;

[0215] The obtaining module 1303 can be specifically used for:

[0216] Based on the parameter values ​​of the variational parameters in the current target quantum circuit, run and measure the current target quantum circuit to obtain the final state;

[0217] When the operation of the current target quantum circuit meets the preset termination condition, the current final state is taken as the final quantum state.

[0218] As can be seen, the embodiments of this application first determine the initial quantum state to be prepared based on the structure of the target protein, wherein the target protein includes proteins whose alignment mode needs to be determined; then, using the initial quantum state and the Hamiltonian corresponding to the obtained target protein, a target quantum circuit is constructed; then, the current target quantum circuit is run and measured to obtain a final quantum state containing the alignment mode of the target protein; finally, based on the final quantum state, the optimal alignment mode of the target protein is determined. Due to the superposition and entanglement of qubits, quantum computing has intrinsic parallelism. Therefore, by utilizing the computing power of noisy qubits, the alignment mode of protein structures can be determined more quickly through quantum circuits, thereby efficiently determining a better alignment mode of protein structures.

[0219] This application also provides a storage medium storing a computer program, wherein the computer program is configured to implement the steps in any of the above method embodiments when running.

[0220] Specifically, in this embodiment, the storage medium can be configured to store a computer program for implementing the following steps:

[0221] S201: Based on the structure of the target protein, determine the initial quantum state to be prepared, wherein the target protein includes a protein whose alignment mode is to be determined;

[0222] S202: Construct the target quantum circuit using the initial quantum state and the Hamiltonian corresponding to the obtained target protein;

[0223] S203: Run and measure the current target quantum circuit to obtain the final quantum state containing the alignment of the target protein;

[0224] S204: Determine the optimal alignment of the target protein based on the final quantum state.

[0225] This application also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the steps in any of the above method embodiments.

[0226] Specifically, the aforementioned electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the aforementioned processor, and the input / output device is connected to the aforementioned processor.

[0227] Specifically, in this embodiment, the processor described above can be configured to implement the following steps via a computer program:

[0228] S201: Based on the structure of the target protein, determine the initial quantum state to be prepared, wherein the target protein includes a protein whose alignment mode is to be determined;

[0229] S202: Construct the target quantum circuit using the initial quantum state and the Hamiltonian corresponding to the obtained target protein;

[0230] S203: Run and measure the current target quantum circuit to obtain the final quantum state containing the alignment of the target protein;

[0231] S204: Determine the optimal alignment of the target protein based on the final quantum state.

[0232] The above description, based on the embodiments shown in the drawings, details the structure, features, and effects of this application. The above description is only a preferred embodiment of this application, but this application does not limit the scope of implementation to what is shown in the drawings. Any changes made in accordance with the concept of this application, or modifications to equivalent embodiments, that do not exceed the spirit covered by the specification and drawings, should be within the protection scope of this application.

Claims

1. A method for determining protein structure alignment, characterized in that, The method includes: Based on the structure of the target protein, all alignment methods are determined, and all alignment methods are encoded to obtain the initial quantum state to be prepared. The target protein includes two proteins with alignment methods to be determined. Construct an initial state preparation circuit for preparing the initial quantum state; construct a simulation circuit for simulating the evolution of the Hamiltonian corresponding to the target protein; construct an adjustment circuit for adjusting the quantum state probability; combine the initial state preparation circuit, the simulation circuit, and the adjustment circuit to obtain the target quantum circuit, wherein the Hamiltonian includes a single-body term that sums over all alignments and a two-body action term that penalizes illegal alignments; Run and measure the current target quantum circuit to obtain the final quantum state containing the alignment of the target protein; Based on the final quantum state, the optimal alignment of the target protein is determined.

2. The method according to claim 1, characterized in that, Based on the structure of the target protein, all alignment methods are determined, and all alignment methods are encoded to obtain the initial quantum state to be prepared, including: Two contact sets corresponding to the target protein are obtained, wherein each contact set contains at least one pair of contact elements, and the pair of contact elements consists of two amino acids in the target protein that satisfy a preset contact relationship; By using two contact sets, all alignment methods are determined, and all alignment methods are encoded to obtain the initial quantum state to be prepared.

3. The method according to claim 2, characterized in that, The process of using two contact sets to determine all alignment methods, encoding all alignment methods, and obtaining the initial quantum state to be prepared includes: Based on the two contact sets, the number of qubits required to prepare the initial quantum state is determined; The determined qubits are grouped using a contact set; For each group, determine the quantum state to be prepared. One contact corresponds to one group, and each quantum state to be prepared includes all alignment methods corresponding to that contact. Based on all the quantum states to be prepared, the initial quantum state is obtained.

4. The method according to claim 3, characterized in that, The determination of the number of qubits required to prepare the initial quantum state based on two contact sets includes: Based on the number of the two contact sets, the number of qubits required to prepare the initial quantum state is determined by the following formula: in, For the required number of qubits, The number of contacts in a contact set. The number of contacts in another contact set. ; The method of grouping the determined number of qubits using one of the contact sets includes: The determined qubits are arranged into groups containing The qubits are grouped together.

5. The method according to claim 3, characterized in that, The Hamiltonian is expressed by the following formula: in, Let this be the Hamiltonian. , Corresponding to the first The and the first One quantum bit, This is the penalty coefficient.

6. The method according to claim 5, characterized in that, The simulated circuit includes variational parameters; The process of running and measuring the current target quantum circuit to obtain the final quantum state containing the alignment of the target protein includes: Based on the parameter values ​​of the variational parameters in the current target quantum circuit, run and measure the current target quantum circuit to obtain the final state; When the operation of the current target quantum circuit meets the preset termination condition, the current final state is taken as the final quantum state.

7. A device for determining protein structure alignment, characterized in that, The device includes: The first determining module is used to determine all alignment methods based on the structure of the target protein, encode all alignment methods, and obtain the initial quantum state to be prepared, wherein the target protein includes two proteins with alignment methods to be determined; A construction module is used to construct an initial state preparation circuit for preparing the initial quantum state; to construct a simulation circuit for simulating the evolution of the Hamiltonian corresponding to the target protein; to construct an adjustment circuit for adjusting the quantum state probability; and to combine the initial state preparation circuit, the simulation circuit, and the adjustment circuit to obtain the target quantum circuit, wherein the Hamiltonian includes a single-body term that sums over all alignments and a two-body action term that penalizes illegal alignments. The module is used to run and measure the current target quantum circuit to obtain the final quantum state containing the alignment of the target protein; The second determining module is used to determine the optimal alignment of the target protein based on the final quantum state.

8. A storage medium, characterized in that, The storage medium stores a computer program, wherein the computer program is configured to implement the method described in any one of claims 1 to 6 when it is run.

9. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to implement the method according to any one of claims 1 to 6.

Images