A protein side chain structure prediction method and device based on a coherent Ising machine

Abstract

The application provides a protein side chain structure prediction method and device based on a coherent Ising machine, and the method comprises the following steps: firstly, obtaining a protein file; then, generating a side chain conformation relationship network graph according to the protein file, wherein the network graph comprises a plurality of side chain conformation nodes and a connection edge established between the side chain conformation nodes and satisfying a predetermined condition; then, searching a solution space of the side chain conformation relationship graph represented by a coherent Ising machine system, and solving a ground state of a Hamiltonian H; and finally, generating a target side chain structure of the protein according to the ground state of the Hamiltonian H. The protein side chain structure prediction method provided by the application can systematically search all possible side chain conformation combinations, ensure that the found solution is globally optimal, and has higher solving accuracy. In addition, the coherent Ising machine system is used, and the calculation efficiency is higher when a large-scale problem is processed.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of computational biology, and particularly relates to a protein side-chain structure prediction method and device based on a coherent Ising machine. BACKGROUND

[0002] The Protein Side-chain Packing Problem (PSPP) is a key challenge in the field of computational biology and structure prediction. The PSPP problem involves predicting the optimal arrangement of amino acid side chains in three-dimensional space, given the known protein backbone structure, to achieve the stability and functionality of the entire protein structure. This problem is recognized as NP-hard due to its complex combinatorial nature and vast search space, meaning that there is no known polynomial-time algorithm that can solve all instances exactly.

[0003] Existing solutions to the PSPP problem include:

[0004] 1. Dead-end Elimination (DEE):

[0005] DEE is a deterministic algorithm that reduces the search space by iteratively eliminating side-chain conformations that cannot belong to the global minimum energy conformation. This method is based on a simple criterion: if a side-chain conformation cannot exist in any possible global minimum energy conformation, it can be removed from the search space. DEE methods can guarantee finding the global optimal solution, but may encounter convergence problems when dealing with large proteins and require huge computational resources, limiting their application on large-scale problems.

[0006] 2. Genetic Algorithms (GA):

[0007] Genetic algorithms are heuristic algorithms that mimic the process of natural selection. In the PSPP problem, genetic algorithms initialize a population of side-chain conformations and then iteratively improve these conformations through selection, crossover, and mutation operations. Eventually, the algorithm converges to a set of low-energy side-chain conformations. Genetic algorithms have the advantage of parallel processing, but may not guarantee finding the global optimal solution, and their performance is highly dependent on parameter settings, which may not be computationally efficient.

[0008] 3. Monte Carlo Methods:

[0009] Monte Carlo method is a calculation method based on random sampling, which is used to solve complex optimization problems. In the PSPP problem, Monte Carlo method selects side chain conformations randomly and calculates their energies, then accepts or rejects new conformations according to energy evaluation. This method is suitable for handling large-scale problems, but its characteristics based on random sampling cannot guarantee to find the global optimal solution, and it may require a large number of samples, high computational cost and slow convergence speed.

[0010] It can be seen that in the PSPP problem, these algorithms can provide approximate solutions to a certain extent, but they are limited in terms of computational efficiency, solution accuracy and scalability. In addition, these methods often require huge computing resources when dealing with large-scale problems, which limits their feasibility in practical applications. SUMMARY

[0011] To solve the above problems, the present application provides a protein side chain structure prediction method and device based on a coherent Ising machine.

[0012] According to a first aspect, a protein side chain structure prediction method is provided, comprising the following steps:

[0013] S1, obtaining a protein file, which includes the spatial coordinates of a plurality of atoms in the protein, and the plurality of atoms form a plurality of amino acid residues.

[0014] S2, generating a side chain conformation relationship network graph according to the protein file, which includes a plurality of side chain conformation nodes and connection edges established between the side chain conformation nodes satisfying a predetermined condition.

[0015] Wherein, the determination of the plurality of side chain conformation nodes comprises: for each amino acid residue in the plurality of amino acid residues, a certain number of side chain conformations are generated by sampling the side chain search space, and the side chain conformations that do not overlap with the main chain of the protein in space are selected as nodes.

[0016] The predetermined condition to be met for establishing the connection edge includes that the two side chain conformation nodes do not belong to the same amino acid residue and do not overlap in space.

[0017] S3, using a coherent Ising machine system to search the solution space represented by the side chain conformation relationship graph, and solving the ground state of Hamiltonian H, wherein the Hamiltonian H is:

[0018]

[0019] Wherein, represents whether the node corresponding to the th side chain conformation of the th amino acid residue is selected, is The weight, Representing the The first amino acid residue The first sidechain conformation and the first The first amino acid residue Whether there are connecting edges between the nodes corresponding to each sidechain construct, K1 is... The coefficient.

[0020] S4. Generate the target side chain structure of the protein based on the ground state of the Hamiltonian H.

[0021] In some embodiments, screening out side chain conformations that do not spatially overlap with the main chain of the protein specifically includes: for each side chain conformation, measuring its minimum distance from the main chain; if the distance is greater than a first threshold, then selecting this amino acid residue side chain conformation as a node.

[0022] In some embodiments, to satisfy the predetermined conditions, the establishment of the connection edge specifically includes: for each pair of side chain conformations that do not belong to the same amino acid residue, measuring the minimum atomic distance between them; if the distance is greater than a second threshold, then establishing a connection edge between their corresponding vertices.

[0023] In some embodiments, the search of the solution space represented by the sidechain conformation diagram using a coherent Ising machine system to solve for the ground state of the Hamiltonian H specifically includes:

[0024] The coherent Ising machine is initialized based on the Hamiltonian H, and the evolution process of the coherent Ising machine system is initiated.

[0025] During the evolution of the coherent Ising machine system, the phase and intensity of the optical field within the coherent Ising machine system are measured multiple times. Based on the measurement results, the parameters of the coherent Ising machine system are adjusted in real time, ultimately obtaining a collective oscillation mode that is much higher than the threshold. The quantum spin state in this mode corresponds to the ground state of the Hamiltonian H.

[0026] In some embodiments, the coherent Ising machine system includes:

[0027] Optical parametric oscillator network for simulating spin interactions in the Ising model;

[0028] A phase-sensitive amplifier is used to amplify signals of a specific phase.

[0029] Field-programmable gate arrays (FPGAs) are used to control optical parametric oscillator networks.

[0030] Phase / intensity measuring device, used to measure the phase and intensity of the optical field inside the cavity;

[0031] optical modulators for changing the phase difference between light beams to simulate the interaction strength between different spin states;

[0032] beam splitters for generating interference effects to simulate the connection between spins in the Ising model;

[0033] optical fibers for serving as transmission media for optical signals.

[0034] According to a second aspect, a protein side chain structure prediction device based on a coherent Ising machine is provided, comprising:

[0035] An acquisition module is configured to acquire a protein file, wherein the protein file comprises spatial coordinates of a plurality of atoms in a protein, and the plurality of atoms form a plurality of amino acid residues.

[0036] A graph generation module is configured to generate a side chain conformation relationship network graph based on the protein file, wherein the side chain conformation relationship network graph comprises a plurality of side chain conformation nodes and connection edges between the side chain conformation nodes that satisfy a predetermined condition. The determination of the plurality of side chain conformation nodes comprises: for each of the plurality of amino acid residues, generating a specific number of side chain conformations by sampling a side chain search space, and selecting side chain conformations that do not overlap with the main chain of the protein in space as nodes from the specific number of side chain conformations; and the predetermined condition comprises: two side chain conformation nodes do not belong to the same amino acid residue and do not overlap in space.

[0037] A solving module is configured to search a solution space represented by the side chain conformation relationship graph using a coherent Ising machine system, and solve a ground state of a Hamiltonian H, wherein the Hamiltonian H is:

[0038]

[0039] wherein, represents whether a node corresponding to a side chain conformation of an i-th amino acid residue is selected, is a weight of the i-th amino acid residue, represents whether there is a connection edge between nodes corresponding to a side chain conformation of an i-th amino acid residue and a side chain conformation of a j-th amino acid residue, and K1 is a coefficient of the i-th amino acid residue.

[0040] An output module is configured to output a target side chain structure of the protein according to the ground state of the Hamiltonian H.

[0041] In some embodiments, the side chain conformations that do not overlap with the main chain of the protein in space are specifically selected by:​​​​​​​​

[0042] For each side chain conformation, measure the minimum distance to the main chain, if the distance is greater than the first threshold, select the amino acid residue side chain conformation as a node;

[0043] In some embodiments, the establishment of the connection edge includes:

[0044] For each pair of side chain conformations not belonging to the same amino acid residue, measure the minimum atomic distance between them, if the distance is greater than the second threshold, establish a connection edge between the vertices corresponding to them.

[0045] In some embodiments, the search of the solution space represented by the side chain conformation relationship graph by using the coherent Ising machine system, and the solving of the ground state of the Hamiltonian H specifically includes:

[0046] According to the Hamiltonian H, initialize the coherent Ising machine, and start the coherent Ising machine system evolution process;

[0047] During the coherent Ising machine system evolution process, the phase and intensity of the light field in the coherent Ising machine system are measured multiple times, and the coherent Ising machine system parameters are adjusted in real time according to the measurement results, and finally a collective oscillation mode much higher than the threshold is obtained, and the quantum spin state under the mode corresponds to the ground state of the Hamiltonian H.

[0048] The protein side chain structure prediction method provided by the application can systematically search all possible side chain conformation combinations, ensure that the found solution is globally optimal, and has higher solving accuracy. In addition, the application uses a coherent Ising machine system, which has higher calculation efficiency when dealing with large-scale problems. The application uses a coherent Ising machine to predict protein side chain folding, improves the traditional solving complexity, and has the advantage of polynomial improvement or constant pre-factor of algorithm complexity, which makes the advantage of the method more and more obvious as the system size increases. BRIEF DESCRIPTION OF DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the application, the following briefly introduces the drawings needed in the embodiment description. Obviously, the drawings in the following description only some embodiments of the application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of these drawings.

[0050] Figure 1 A protein side chain structure prediction method flowchart provided by the application;

[0051] Figure 2 A coherent Ising machine system structure schematic diagram provided by the embodiment of the application;

[0052] Figure 3A protein side chain structure prediction device provided by the present application is shown in a schematic diagram. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below with reference to the drawings.

[0054] The Coherent Ising Machine (CIM) is a computing platform based on quantum optics simulation of the Ising model, which has the advantages of fast computing speed and strong scalability, and is suitable for processing various large-scale combinatorial optimization problems, especially in processing large-scale quantum bits and NP-hard problems. In processing these problems, the CIM can ensure that the solution found is globally optimal while maintaining the efficiency of the computation and the accuracy of the solution.

[0055] The CIM provides a new solution to the PSPP problem with its unique parallel processing capability and high-speed computing performance. Based on this, the present application first converts the PSPP problem into an Ising model optimization problem by designing the Hamiltonian of the Coherent Ising Machine, so that it can be efficiently computed. Then, the Coherent Ising Machine is used to solve the PSPP problem, ensuring that the solution found is globally optimal.

[0056] The protein side chain structure prediction method provided by the present application will be described below with reference to the drawings. The flowchart of the method is shown in Figure 1 The method comprises the following steps:

[0057] S1, obtaining a protein file, which includes the spatial coordinates of a plurality of atoms in the protein, the plurality of atoms forming a plurality of amino acid residues.

[0058] It can be understood that a protein is a molecular chain formed by a plurality of amino acid residues connected by peptide bonds, and each amino acid residue contains a backbone and a side chain. The backbone is composed of a repeating sequence of atoms (including nitrogen atoms, alpha carbon atoms and carbonyl carbon atoms), forming the skeletal structure of the protein. The side chain is a unique chemical group connected to the alpha carbon atom of the backbone, composed of different types of atoms, determining the characteristics and functions of the amino acid. All these structures are ultimately formed by the arrangement and bonding of atoms, and the atom is the most basic unit of the protein.

[0059] S2, generating a side chain conformation relationship network diagram according to the protein file, which includes a plurality of side chain conformation nodes and connection edges between the side chain conformation nodes that satisfy a predetermined condition.

[0060] The determining of the plurality of side chain conformation nodes comprises: for each of the plurality of amino acid residues, generating a specific number of side chain conformations by sampling a side chain search space, i.e., sampling a specific number of side chain conformations from the side chain search space of the amino acid residue, and screening out, as nodes, side chain conformations that are spatially non-overlapping with the main chain of the protein. It can be understood that the side chain conformation refers to the spatial arrangement or geometric shape of the side chain of an amino acid in a protein.

[0061] In some embodiments, the screening out of the side chain conformations that are spatially non-overlapping with the main chain of the protein specifically comprises: for each side chain conformation, measuring the minimum distance thereof from the main chain, and if the distance is greater than a first threshold, selecting the side chain conformation of the amino acid residue as a node. It should be understood that the "first" in the "first threshold" and similar terms such as "second" elsewhere in the text are used to distinguish the same things, and do not have other limitations such as ordering.

[0062] The predetermined condition to be met for establishing a connection edge comprises that two side chain conformation nodes do not belong to the same amino acid residue and are spatially non-overlapping.

[0063] In some embodiments, to meet the predetermined condition, the establishment of the connection edge specifically comprises: for each pair of side chain conformations that do not belong to the same amino acid residue, measuring the minimum atomic distance therebetween, and if the distance is greater than a second threshold, establishing a connection edge between the vertices corresponding to the two side chain conformations.

[0064] Through the above steps, all possible side chain conformations of the amino acid residues and their relationships are converted into a side chain conformation relationship network graph G=(V, E), where V is a node set and E is an edge set. V contains nodes that do not collide with the main chain (i.e., the th conformation of the th residue), and E contains { , } that meet the predetermined condition (i.e., the th conformation of the th residue and the th conformation of the th residue are spatially non-overlapping). In this way, the protein side chain folding problem is converted into a graph theory problem. Next, the best arrangement of all amino acid residue side chains is found by finding the maximum clique in the graph G, which can be achieved by constructing the Hamiltonian H of the graph G and finding the ground state of the Hamiltonian H by using the CIM, and the specific steps are as follows:

[0065] S3, using a coherent Ising machine system, searching the solution space represented by the side chain conformation relationship graph, and solving the ground state of the Hamiltonian H, wherein the Hamiltonian H is:

[0066] (1)

[0067] The meanings and possible values ​​of each parameter are as follows:

[0068] Representing the The first amino acid residue Whether a node corresponding to a sidechain conformation is selected corresponds to the node spin state in a coherent Ising machine. For example, a spin of +1 can be set to indicate that this conformation is selected.

[0069] for The weight can be set to 1.

[0070] Representing the The first amino acid residue The first sidechain conformation and the first The first amino acid residue Does each sidechain conformation have connecting edges between its corresponding nodes? For example, a value of 1 is used when there is a connecting edge, and a value of 0 is used when there is no connecting edge.

[0071] K1 is The coefficient can be set to any positive number, such as 1.

[0072] In this step, the ground state of Hamiltonian H shown in equation (1) can be found using the coherent Ising machine system, which is also the largest clique of graph G.

[0073] In some embodiments, the search of the solution space represented by the sidechain conformation diagram using a coherent Ising machine system to solve for the ground state of the Hamiltonian H specifically includes:

[0074] S31. Initialize the coherent Ising machine according to the Hamiltonian H.

[0075] S32. Initiate the evolution process of the coherent Ising machine system by increasing the pump amplitude from 0 to obtain a collective oscillation mode that is much higher than the threshold. The quantum spin state in this mode corresponds to the ground state of the Hamiltonian H.

[0076] It should be understood that during the evolution of the coherent Ising machine system, the phase and intensity of the optical field within the coherent Ising machine system can be measured multiple times, and the parameters of the coherent Ising machine system can be adjusted in real time based on the measurement results.

[0077] From the above, the ground state of the Hamiltonian H can be obtained using the coherent Ising machine system.

[0078] S4. Generate the target side chain structure of the protein based on the ground state of the Hamiltonian H.

[0079] Specifically, the ground state of the Hamiltonian H, or the nodes in the largest cluster, are inverse-encoded into protein residue side chain conformations as the final prediction results.

[0080] Exemplarily, the output form of the final prediction results can be a protein three-dimensional structure diagram or protein atomic spatial coordinates.

[0081] The above is the protein side chain structure prediction method based on the coherent Ising machine provided by the present application, and the coherent Ising machine system used in the method will be briefly introduced below.

[0082] The coherent Ising machine system uses a doubly resonant optical parametric oscillator (DOPO) to realize artificial spins, and realizes the enhancement of light signals with specific phases by placing a phase sensitive amplifier (PSA) in the optical cavity. The PSA is an optical amplifier based on optical parametric amplification, which can effectively amplify the 0 and π phase components relative to the pump phase. Therefore, the DOPO only uses 0 or π phase above the oscillation threshold; therefore, discrete phase states can be used to represent Ising spin states. The interaction between DOPO pulses is realized using a measurement feedback technique, which repeatedly measures the feedback modulation process in the cavity, while increasing the pump amplitude from 0, and finally obtains a "strongest" collective oscillation mode much higher than the threshold, which corresponds to the best solution of a given Ising problem.

[0083] The schematic diagram of the coherent Ising machine system is shown in Figure 2 , which includes:

[0084] (1) Doubly resonant optical parametric oscillator network for simulating spin interactions in the Ising model.

[0085] (2) Phase sensitive amplifier for amplifying signals with specific phases.

[0086] (3) Field-programmable gate array (FPGA) for controlling the doubly resonant optical parametric oscillator network.

[0087] The FPGA can be programmed to configure its internal circuit, which can be used to adjust the parameters of the coherent Ising machine system in real time according to the measurement results during the evolution of the coherent Ising machine system.

[0088] (4) Phase / intensity measurer for measuring the phase and intensity of the light field in the cavity.

[0089] (5) Optical modulator for changing the phase difference between light beams to simulate the interaction strength between different spin states.

[0090] (6) a beam splitter for generating an interference effect, simulating the connection relationship between spins in the Ising model.

[0091] (7) an optical fiber used as a transmission medium for optical signals.

[0092] As can be seen from the above, compared with the prior art, the protein side chain structure prediction method provided by the present application has the following beneficial effects:

[0093] 1. Guarantee of global optimal solution: The method can systematically search all possible side chain conformation combinations, ensuring that the solution found is globally optimal. This is particularly important in the NP-hard PSPP problem, as many heuristic algorithms can quickly find approximate solutions, but cannot guarantee the optimality of the solution.

[0094] 2. High precision prediction ability: Compared with traditional heuristic or approximate algorithms, the method has higher solution accuracy and can provide more accurate results, meeting the accurate solution obtained by the enumeration algorithm.

[0095] 3. Acceleration potential: The method is implemented based on a coherent Ising machine system, providing higher approximate solution accuracy than a classical computer, and the advantage becomes more and more obvious as the problem size increases. The acceleration advantage of the Ising machine is beyond the field of complexity theory, providing a polynomial improvement in scaling or constant pre-factor advantage, which has a great impact on the running time of large problems, i.e. the advantage of the method becomes more and more obvious as the system size increases.

[0096] 4. Ability to directly obtain three-dimensional conformation: The method can directly obtain the three-dimensional conformation of the protein side chain folding, providing a more convenient and fast solution for the biopharmaceutical field.

[0097] The present application also provides a protein side chain structure prediction device 300 based on a coherent Ising machine, a schematic diagram of which is shown in Figure 3 , comprising:

[0098] An acquisition module 301 is configured to acquire a protein file, which includes the spatial coordinates of a plurality of atoms in the protein, and the plurality of atoms form a plurality of amino acid residues.

[0099] A graph generation module 302 is configured to generate a side chain conformation relationship network graph according to the protein file, which includes a plurality of side chain conformation nodes and connection edges established between the side chain conformation nodes that satisfy a predetermined condition. The determination of the plurality of side chain conformation nodes includes: for each amino acid residue in the plurality of amino acid residues, generating a specific number of side chain conformations by sampling the side chain search space, and selecting the side chain conformations that do not overlap with the main chain of the protein in space as nodes from the side chain conformations. The predetermined condition includes that two side chain conformation nodes do not belong to the same amino acid residue and do not overlap in space.

[0100] a solving module 303, configured to search a solution space of the side chain conformation relationship graph representation by using a coherent Ising machine system, and solve a ground state of a Hamiltonian H, the Hamiltonian H being:

[0101]

[0102] wherein, represents whether a node corresponding to a k-th side chain conformation of a j-th amino acid residue is selected, is a weight of the k-th side chain conformation of the j-th amino acid residue, represents whether there is a connecting edge between nodes corresponding to a k-th side chain conformation of a j-th amino acid residue and a k'-th side chain conformation of a i-th amino acid residue, and K1 is a coefficient of the k-th side chain conformation of the j-th amino acid residue.

[0103] An output module 304 is configured to output a target side chain structure of the protein according to the ground state of the Hamiltonian H.

[0104] It should be noted that the apparatus described above can perform the protein side chain structure prediction method described above, and the functions of each module can be referred to the description of the method, and will not be described herein.

[0105] In the description of the embodiments of the present application, the words "exemplary", "for example", or "for instance" are used to mean serving as an example, instance, or illustration. Any embodiment or design described as "exemplary", "for example", or "for instance" in the embodiments of the present application should not be construed as being more advantageous or superior than other embodiments or designs. In fact, the use of "exemplary", "for example", or "for instance" is intended to present concepts in a concrete manner.

[0106] In the description of the embodiments of the present application, the term "and / or" is merely used to describe an associated relationship with associated objects, and means that there can be three relationships, for example, A and / or B can mean that there are three cases of A alone, B alone, and A and B simultaneously. In addition, unless otherwise specified, the term "multiple" means two or more.

[0107] In addition, the terms "comprising", "including", "having" and their conjugates mean "including but not limited to", unless otherwise specifically emphasized.

[0108] ​​​​​​​​The above detailed description of the specific embodiments of the present application has been given to illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made on the basis of the technical solutions of the present application shall be included in the protection scope of the present application.

Claims

1. A method of predicting a side chain structure of a protein, characterized by, The method comprises the following steps: obtaining a protein file comprising spatial coordinates of a plurality of atoms in a protein, the plurality of atoms forming a plurality of amino acid residues; generating a side chain conformation relationship network graph according to the protein file, the side chain conformation relationship network graph comprising a plurality of side chain conformation nodes and connection edges between the side chain conformation nodes satisfying a predetermined condition; the determination of the plurality of side chain conformation nodes comprises: for each of the plurality of amino acid residues, generating a specific number of side chain conformations by sampling a side chain search space, and selecting side chain conformations that are spatially non-overlapping with the main chain of the protein as nodes; and the predetermined condition comprises: two side chain conformation nodes do not belong to the same amino acid residue and are spatially non-overlapping; using a coherent Ising machine system to search the solution space represented by the side chain conformation relationship graph, and solving the ground state of a Hamiltonian H, the Hamiltonian H being: in, Representing the The first amino acid residue Whether the node corresponding to each sidechain conformation is selected. for The weight, Representing the The first amino acid residue The first sidechain conformation and the first The first amino acid residue Whether there are connecting edges between the nodes corresponding to each sidechain construct, K1 is... The coefficient; generating a target side chain structure of the protein according to the ground state of the Hamiltonian H; the using of the coherent Ising machine system to search the solution space represented by the side chain conformation relationship graph and solve the ground state of the Hamiltonian H specifically comprises: initializing the coherent Ising machine according to the Hamiltonian H, and starting the coherent Ising machine system evolution process; during the coherent Ising machine system evolution process, measuring the phase and intensity of the light field in the coherent Ising machine system multiple times, and adjusting the parameters of the coherent Ising machine system in real time according to the measurement results, and finally obtaining the collective oscillation mode of the quantum spin state corresponding to the ground state of the Hamiltonian H.

2. The method of claim 1, wherein, the selection of the side chain conformations that are spatially non-overlapping with the main chain of the protein specifically comprises: for each side chain conformation, measuring the minimum distance between the side chain conformation and the main chain, and if the distance is greater than a first threshold, selecting the side chain conformation of the amino acid residue as a node.

3. The method of claim 1, wherein, the establishment of the connection edges comprises: for each pair of side chain conformations that do not belong to the same amino acid residue, measuring the minimum atomic distance between the side chain conformations, and if the distance is greater than a second threshold, establishing a connection edge between the vertices corresponding to the side chain conformations.

4. The method of claim 1, wherein, the coherent Ising machine system comprises: an optical parametric oscillator network for simulating spin interactions in the Ising model; a phase-sensitive amplifier for amplifying signals of a specific phase; a field programmable gate array for controlling the optical parametric oscillator network; a phase / intensity measurer for measuring the phase and intensity of the light field in the cavity; an optical modulator for changing the phase difference between light beams to simulate the interaction strength between different spin states; a beam splitter for generating interference effects to simulate the connection relationship between spins in the Ising model; optical fibers serving as transmission media for optical signals.

5. A protein side chain structure prediction apparatus based on a coherent Ising machine, characterized by, The method comprises the following steps: obtaining a protein file comprising spatial coordinates of a plurality of atoms in a protein, the plurality of atoms forming a plurality of amino acid residues; generating a side chain conformation relationship network graph according to the protein file, the side chain conformation relationship network graph comprising a plurality of side chain conformation nodes and connection edges between the side chain conformation nodes satisfying a predetermined condition; The determination of the plurality of side chain conformation nodes comprises: for each of the plurality of amino acid residues, generating a specific number of side chain conformations by sampling a side chain search space, and screening side chain conformations that are spatially non-overlapping with the main chain of the protein as nodes; and the predetermined condition comprises: two side chain conformation nodes do not belong to the same amino acid residue and are spatially non-overlapping. The solving module is configured to search, by using a coherent Ising machine system, a solution space represented by the side chain conformation graph, to solve a ground state of a Hamiltonian H, the Hamiltonian H being: in, Representing the The first amino acid residue Whether the node corresponding to each sidechain conformation is selected. for The weight, Representing the The first amino acid residue The first sidechain conformation and the first The first amino acid residue Whether there are connecting edges between the nodes corresponding to each sidechain construct, K1 is... The coefficient; The searching, by using the coherent Ising machine system, the solution space represented by the side chain conformation graph, to solve the ground state of the Hamiltonian H specifically comprises: According to the Hamiltonian H, initializing the coherent Ising machine, and starting the coherent Ising machine system evolution process; During the coherent Ising machine system evolution process, measuring the phase and intensity of the optical field in the coherent Ising machine system multiple times, and adjusting the parameters of the coherent Ising machine system in real time according to the measurement results, to finally obtain a collective oscillation mode corresponding to the ground state of the Hamiltonian H of the quantum spin state; The output module is configured to output the target side chain structure of the protein according to the ground state of the Hamiltonian H.

6. The apparatus of claim 5, wherein, The screening of the side chain conformation that is spatially non-overlapping with the main chain of the protein specifically comprises: For each side chain conformation, measuring the minimum distance thereof from the main chain, and if the distance is greater than a first threshold value, selecting the side chain conformation of the amino acid residue as a node.

7. The apparatus of claim 5, wherein, The establishment of the connection edge comprises: For each pair of side chain conformations that do not belong to the same amino acid residue, measuring the minimum atomic distance therebetween, and if the distance is greater than a second threshold value, establishing a connection edge between the vertices corresponding to the side chain conformations.

8. The apparatus of claim 5, wherein, The coherent Ising machine system comprises: An optical parametric oscillator network for simulating spin interactions in the Ising model; A phase-sensitive amplifier for amplifying signals of a specific phase; A field programmable gate array for controlling the optical parametric oscillator network; A phase / intensity measurer for measuring the phase and intensity of the optical field in the cavity; An optical modulator for changing the phase difference between the light beams, thereby simulating the interaction strength between different spin states; A beam splitter for generating interference effects to simulate the connection relationship between spins in the Ising model; An optical fiber serving as a transmission medium for optical signals.

Images